fix controlnets for latest diffusers v

QWEN_IMAGE_IMPLEMENTATION.md

@@ -1,128 +0,0 @@
-# Qwen-Image Implementation for InvokeAI
-
-## Overview
-
-This implementation adds support for the Qwen-Image family of models to InvokeAI. Qwen-Image is a 20B parameter Multimodal Diffusion Transformer (MMDiT) model that excels at complex text rendering and precise image editing.
-
-## Model Setup
-
-### 1. Download the Qwen-Image Model
-```bash
-# Option 1: Using git (recommended for large models)
-git clone https://huggingface.co/Qwen/Qwen-Image invokeai/models/qwen-image/Qwen-Image
-
-# Option 2: Using huggingface-cli
-huggingface-cli download Qwen/Qwen-Image --local-dir invokeai/models/qwen-image/Qwen-Image
-```
-
-### 2. Download Qwen2.5-VL Text Encoder
-Qwen-Image uses Qwen2.5-VL-7B as its text encoder (not CLIP):
-```bash
-git clone https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct invokeai/models/qwen-image/Qwen2.5-VL-7B-Instruct
-```
-
-## Model Architecture
-
-### Components
-1. **Transformer**: QwenImageTransformer2DModel (MMDiT architecture, 20B parameters)
-2. **Text Encoder**: Qwen2.5-VL-7B-Instruct (7B parameter vision-language model)
-3. **VAE**: AutoencoderKLQwenImage (bundled with main model in `/vae` subdirectory)
-4. **Scheduler**: FlowMatchEulerDiscreteScheduler
-
-### Key Features
-- **Complex Text Rendering**: Superior ability to render text accurately in images
-- **Bundled VAE**: The model includes its own custom VAE (no separate download needed)
-- **Large Text Encoder**: Uses a 7B parameter VLM instead of traditional CLIP
-- **Optional VAE Override**: Can use custom VAE models if desired
-
-## Components Implemented
-
-### Backend Components
-1. **Model Taxonomy** (`taxonomy.py`): Added `QwenImage = "qwen-image"` base model type
-2. **Model Configuration** (`config.py`): Uses MainDiffusersConfig for Qwen-Image models
-3. **Model Loader** (`qwen_image.py`): Loads models and submodels via diffusers
-4. **Model Loader Node** (`qwen_image_model_loader.py`): Loads transformer, text encoder, and VAE
-5. **Text Encoder Node** (`qwen_image_text_encoder.py`): Encodes prompts using Qwen2.5-VL
-6. **Denoising Node** (`qwen_image_denoise.py`): Generates images using QwenImagePipeline
-
-### Frontend Components
-1. **UI Types**: Added QwenImageMainModel, Qwen2_5VLModel field types
-2. **Field Components**: Created input components for model selection
-3. **Type Guards**: Added model detection and filtering functions
-4. **Hooks**: Model loading hooks for UI dropdowns
-
-## Dependencies Updated
-
-- Updated `pyproject.toml` to use `diffusers[torch]==0.35.0` (from 0.33.0) to support Qwen-Image models
-
-## Usage in InvokeAI
-
-### Node Graph Setup
-1. Add a **"Main Model - Qwen-Image"** loader node
-2. Select your Qwen-Image model from the dropdown
-3. Select the Qwen2.5-VL model for text encoding
-4. VAE field is optional (uses bundled VAE if left empty)
-5. Connect to **Qwen-Image Text Encoder** node
-6. Connect to **Qwen-Image Denoise** node
-7. Add **VAE Decode** node to convert latents to images
-
-### Model Selection
-- **Main Model**: Select from models with base type "qwen-image"
-- **Text Encoder**: Select Qwen2.5-VL-7B-Instruct
-- **VAE**: Optional - leave empty to use bundled VAE, or select a custom VAE
-
-## Troubleshooting
-
-### VAE Not Showing Up
-The Qwen-Image VAE is bundled with the main model. You don't need to download or select a separate VAE - just leave the VAE field empty to use the bundled one.
-
-### Memory Issues
-Qwen-Image is a large model (20B parameters) and Qwen2.5-VL is 7B parameters. Together they require significant resources:
-
-**Memory Requirements:**
-- **Minimum**: 24GB VRAM (with optimizations)
-- **Recommended**: 32GB+ VRAM for smooth operation
-- **System RAM**: 32GB+ recommended
-
-**Optimization Tips:**
-1. **Use bfloat16 precision**: Reduces memory by ~50%
-   ```python
-   torch_dtype=torch.bfloat16
-   ```
-
-2. **Enable CPU offloading**: Move unused models to system RAM
-   - InvokeAI's model manager handles this automatically when configured
-
-3. **Use quantized versions**: 
-   - Try `diffusers/qwen-image-nf4` for 4-bit quantization
-   - Reduces memory usage by ~75% with minimal quality loss
-
-4. **Adjust cache settings**: In InvokeAI settings:
-   - Reduce `ram_cache_size` if running out of system RAM
-   - Reduce `vram_cache_size` if getting CUDA OOM errors
-
-5. **Load models sequentially**: Don't load all models at once
-   - The model manager now properly calculates sizes for better memory management
-
-### Model Not Loading
-- Ensure the model is in the correct directory structure
-- Check that both Qwen-Image and Qwen2.5-VL models are downloaded
-- Verify diffusers version is 0.35.0 or higher
-
-## Future Enhancements
-
-1. **Image Editing**: Support for Qwen-Image-Edit variant
-2. **LoRA Support**: Fine-tuning capabilities
-3. **Optimizations**: Quantization and speed improvements (Qwen-Image-Lightning)
-4. **Advanced Features**: Image-to-image, inpainting, controlnet support
-
-## Files Modified/Created
-
-- `/invokeai/backend/model_manager/taxonomy.py` (modified)
-- `/invokeai/backend/model_manager/config.py` (modified)
-- `/invokeai/backend/model_manager/load/model_loaders/qwen_image.py` (created)
-- `/invokeai/app/invocations/fields.py` (modified)
-- `/invokeai/app/invocations/primitives.py` (modified)
-- `/invokeai/app/invocations/qwen_image_text_encoder.py` (created)
-- `/invokeai/app/invocations/qwen_image_denoise.py` (created)
-- `/pyproject.toml` (modified)

invokeai/app/invocations/cogview4_model_loader.py

@@ -5,7 +5,7 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
 from invokeai.app.invocations.model import (
     GlmEncoderField,
     ModelIdentifierField,
@@ -14,7 +14,6 @@ from invokeai.app.invocations.model import (
 )
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.backend.model_manager.config import SubModelType
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType
 
 
 @invocation_output("cogview4_model_loader_output")
@@ -39,9 +38,8 @@ class CogView4ModelLoaderInvocation(BaseInvocation):
 
     model: ModelIdentifierField = InputField(
         description=FieldDescriptions.cogview4_model,
+        ui_type=UIType.CogView4MainModel,
         input=Input.Direct,
-        ui_model_base=BaseModelType.CogView4,
-        ui_model_type=ModelType.Main,
     )
 
     def invoke(self, context: InvocationContext) -> CogView4ModelLoaderOutput:

invokeai/app/invocations/controlnet.py

@@ -16,6 +16,7 @@ from invokeai.app.invocations.fields import (
     ImageField,
     InputField,
     OutputField,
+    UIType,
 )
 from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.primitives import ImageOutput
@@ -27,7 +28,6 @@ from invokeai.app.util.controlnet_utils import (
     heuristic_resize_fast,
 )
 from invokeai.backend.image_util.util import np_to_pil, pil_to_np
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType
 
 
 class ControlField(BaseModel):
@@ -63,17 +63,13 @@ class ControlOutput(BaseInvocationOutput):
     control: ControlField = OutputField(description=FieldDescriptions.control)
 
 
-@invocation(
-    "controlnet", title="ControlNet - SD1.5, SD2, SDXL", tags=["controlnet"], category="controlnet", version="1.1.3"
-)
+@invocation("controlnet", title="ControlNet - SD1.5, SDXL", tags=["controlnet"], category="controlnet", version="1.1.3")
 class ControlNetInvocation(BaseInvocation):
     """Collects ControlNet info to pass to other nodes"""
 
     image: ImageField = InputField(description="The control image")
     control_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.controlnet_model,
-        ui_model_base=[BaseModelType.StableDiffusion1, BaseModelType.StableDiffusion2, BaseModelType.StableDiffusionXL],
-        ui_model_type=ModelType.ControlNet,
+        description=FieldDescriptions.controlnet_model, ui_type=UIType.ControlNetModel
     )
     control_weight: Union[float, List[float]] = InputField(
         default=1.0, ge=-1, le=2, description="The weight given to the ControlNet"

invokeai/app/invocations/fields.py

@@ -7,13 +7,6 @@ from pydantic_core import PydanticUndefined
 
 from invokeai.app.util.metaenum import MetaEnum
 from invokeai.backend.image_util.segment_anything.shared import BoundingBox
-from invokeai.backend.model_manager.taxonomy import (
-    BaseModelType,
-    ClipVariantType,
-    ModelFormat,
-    ModelType,
-    ModelVariantType,
-)
 from invokeai.backend.util.logging import InvokeAILogger
 
 logger = InvokeAILogger.get_logger()
@@ -46,9 +39,42 @@ class UIType(str, Enum, metaclass=MetaEnum):
     used, and the type will be ignored. They are included here for backwards compatibility.
     """
 
+    # region Model Field Types
+    MainModel = "MainModelField"
+    CogView4MainModel = "CogView4MainModelField"
+    FluxMainModel = "FluxMainModelField"
+    SD3MainModel = "SD3MainModelField"
+    SDXLMainModel = "SDXLMainModelField"
+    SDXLRefinerModel = "SDXLRefinerModelField"
+    ONNXModel = "ONNXModelField"
+    VAEModel = "VAEModelField"
+    FluxVAEModel = "FluxVAEModelField"
+    LoRAModel = "LoRAModelField"
+    ControlNetModel = "ControlNetModelField"
+    IPAdapterModel = "IPAdapterModelField"
+    T2IAdapterModel = "T2IAdapterModelField"
+    T5EncoderModel = "T5EncoderModelField"
+    CLIPEmbedModel = "CLIPEmbedModelField"
+    CLIPLEmbedModel = "CLIPLEmbedModelField"
+    CLIPGEmbedModel = "CLIPGEmbedModelField"
+    SpandrelImageToImageModel = "SpandrelImageToImageModelField"
+    ControlLoRAModel = "ControlLoRAModelField"
+    SigLipModel = "SigLipModelField"
+    FluxReduxModel = "FluxReduxModelField"
+    LlavaOnevisionModel = "LLaVAModelField"
+    Imagen3Model = "Imagen3ModelField"
+    Imagen4Model = "Imagen4ModelField"
+    ChatGPT4oModel = "ChatGPT4oModelField"
+    Gemini2_5Model = "Gemini2_5ModelField"
+    FluxKontextModel = "FluxKontextModelField"
+    Veo3Model = "Veo3ModelField"
+    RunwayModel = "RunwayModelField"
+    # endregion
+
     # region Misc Field Types
     Scheduler = "SchedulerField"
     Any = "AnyField"
+    Video = "VideoField"
     # endregion
 
     # region Internal Field Types
@@ -98,38 +124,6 @@ class UIType(str, Enum, metaclass=MetaEnum):
     MetadataItemPolymorphic = "DEPRECATED_MetadataItemPolymorphic"
     MetadataDict = "DEPRECATED_MetadataDict"
 
-    # Deprecated Model Field Types - use ui_model_[base|type|variant|format] instead
-    MainModel = "DEPRECATED_MainModelField"
-    CogView4MainModel = "DEPRECATED_CogView4MainModelField"
-    FluxMainModel = "DEPRECATED_FluxMainModelField"
-    SD3MainModel = "DEPRECATED_SD3MainModelField"
-    SDXLMainModel = "DEPRECATED_SDXLMainModelField"
-    SDXLRefinerModel = "DEPRECATED_SDXLRefinerModelField"
-    ONNXModel = "DEPRECATED_ONNXModelField"
-    VAEModel = "DEPRECATED_VAEModelField"
-    FluxVAEModel = "DEPRECATED_FluxVAEModelField"
-    LoRAModel = "DEPRECATED_LoRAModelField"
-    ControlNetModel = "DEPRECATED_ControlNetModelField"
-    IPAdapterModel = "DEPRECATED_IPAdapterModelField"
-    T2IAdapterModel = "DEPRECATED_T2IAdapterModelField"
-    T5EncoderModel = "DEPRECATED_T5EncoderModelField"
-    CLIPEmbedModel = "DEPRECATED_CLIPEmbedModelField"
-    CLIPLEmbedModel = "DEPRECATED_CLIPLEmbedModelField"
-    CLIPGEmbedModel = "DEPRECATED_CLIPGEmbedModelField"
-    SpandrelImageToImageModel = "DEPRECATED_SpandrelImageToImageModelField"
-    ControlLoRAModel = "DEPRECATED_ControlLoRAModelField"
-    SigLipModel = "DEPRECATED_SigLipModelField"
-    FluxReduxModel = "DEPRECATED_FluxReduxModelField"
-    LlavaOnevisionModel = "DEPRECATED_LLaVAModelField"
-    Imagen3Model = "DEPRECATED_Imagen3ModelField"
-    Imagen4Model = "DEPRECATED_Imagen4ModelField"
-    ChatGPT4oModel = "DEPRECATED_ChatGPT4oModelField"
-    Gemini2_5Model = "DEPRECATED_Gemini2_5ModelField"
-    FluxKontextModel = "DEPRECATED_FluxKontextModelField"
-    Veo3Model = "DEPRECATED_Veo3ModelField"
-    RunwayModel = "DEPRECATED_RunwayModelField"
-    # endregion
-
 
 class UIComponent(str, Enum, metaclass=MetaEnum):
     """
@@ -327,12 +321,6 @@ class CogView4ConditioningField(BaseModel):
     conditioning_name: str = Field(description="The name of conditioning tensor")
 
 
-class QwenImageConditioningField(BaseModel):
-    """A conditioning tensor primitive value for Qwen-Image"""
-
-    conditioning_name: str = Field(description="The name of conditioning tensor")
-
-
 class ConditioningField(BaseModel):
     """A conditioning tensor primitive value"""
 
@@ -421,10 +409,6 @@ class InputFieldJSONSchemaExtra(BaseModel):
     ui_component: Optional[UIComponent] = None
     ui_order: Optional[int] = None
     ui_choice_labels: Optional[dict[str, str]] = None
-    ui_model_base: Optional[list[BaseModelType]] = None
-    ui_model_type: Optional[list[ModelType]] = None
-    ui_model_variant: Optional[list[ClipVariantType | ModelVariantType]] = None
-    ui_model_format: Optional[list[ModelFormat]] = None
 
     model_config = ConfigDict(
         validate_assignment=True,
@@ -481,9 +465,9 @@ class OutputFieldJSONSchemaExtra(BaseModel):
     """
 
     field_kind: FieldKind
-    ui_hidden: bool = False
-    ui_order: Optional[int] = None
-    ui_type: Optional[UIType] = None
+    ui_hidden: bool
+    ui_type: Optional[UIType]
+    ui_order: Optional[int]
 
     model_config = ConfigDict(
         validate_assignment=True,
@@ -517,63 +501,35 @@ def InputField(
     ui_hidden: Optional[bool] = None,
     ui_order: Optional[int] = None,
     ui_choice_labels: Optional[dict[str, str]] = None,
-    ui_model_base: Optional[BaseModelType | list[BaseModelType]] = None,
-    ui_model_type: Optional[ModelType | list[ModelType]] = None,
-    ui_model_variant: Optional[ClipVariantType | ModelVariantType | list[ClipVariantType | ModelVariantType]] = None,
-    ui_model_format: Optional[ModelFormat | list[ModelFormat]] = None,
 ) -> Any:
     """
     Creates an input field for an invocation.
 
-    This is a wrapper for Pydantic's [Field](https://docs.pydantic.dev/latest/api/fields/#pydantic.fields.Field)
+    This is a wrapper for Pydantic's [Field](https://docs.pydantic.dev/latest/api/fields/#pydantic.fields.Field) \
     that adds a few extra parameters to support graph execution and the node editor UI.
 
-    If the field is a `ModelIdentifierField`, use the `ui_model_[base|type|variant|format]` args to filter the model list
-    in the Workflow Editor. Otherwise, use `ui_type` to provide extra type hints for the UI.
+    :param Input input: [Input.Any] The kind of input this field requires. \
+      `Input.Direct` means a value must be provided on instantiation. \
+      `Input.Connection` means the value must be provided by a connection. \
+      `Input.Any` means either will do.
 
-    Don't use both `ui_type` and `ui_model_[base|type|variant|format]` - if both are provided, a warning will be
-    logged and `ui_type` will be ignored.
+    :param UIType ui_type: [None] Optionally provides an extra type hint for the UI. \
+      In some situations, the field's type is not enough to infer the correct UI type. \
+      For example, model selection fields should render a dropdown UI component to select a model. \
+      Internally, there is no difference between SD-1, SD-2 and SDXL model fields, they all use \
+      `MainModelField`. So to ensure the base-model-specific UI is rendered, you can use \
+      `UIType.SDXLMainModelField` to indicate that the field is an SDXL main model field.
 
-    Args:
-        input: The kind of input this field requires.
-        - `Input.Direct` means a value must be provided on instantiation.
-        - `Input.Connection` means the value must be provided by a connection.
-        - `Input.Any` means either will do.
+    :param UIComponent ui_component: [None] Optionally specifies a specific component to use in the UI. \
+      The UI will always render a suitable component, but sometimes you want something different than the default. \
+      For example, a `string` field will default to a single-line input, but you may want a multi-line textarea instead. \
+      For this case, you could provide `UIComponent.Textarea`.
 
-        ui_type: Optionally provides an extra type hint for the UI. In some situations, the field's type is not enough
-        to infer the correct UI type. For example, Scheduler fields are enums, but we want to render a special scheduler
-        dropdown in the UI. Use `UIType.Scheduler` to indicate this.
+    :param bool ui_hidden: [False] Specifies whether or not this field should be hidden in the UI.
 
-        ui_component: Optionally specifies a specific component to use in the UI. The UI will always render a suitable
-        component, but sometimes you want something different than the default. For example, a `string` field will
-        default to a single-line input, but you may want a multi-line textarea instead. In this case, you could use
-        `UIComponent.Textarea`.
+    :param int ui_order: [None] Specifies the order in which this field should be rendered in the UI.
 
-        ui_hidden: Specifies whether or not this field should be hidden in the UI.
-
-        ui_order: Specifies the order in which this field should be rendered in the UI. If omitted, the field will be
-        rendered after all fields with an explicit order, in the order they are defined in the Invocation class.
-
-        ui_model_base: Specifies the base model architectures to filter the model list by in the Workflow Editor. For
-        example, `ui_model_base=BaseModelType.StableDiffusionXL` will show only SDXL architecture models. This arg is
-        only valid if this Input field is annotated as a `ModelIdentifierField`.
-
-        ui_model_type: Specifies the model type(s) to filter the model list by in the Workflow Editor. For example,
-        `ui_model_type=ModelType.VAE` will show only VAE models. This arg is only valid if this Input field is
-        annotated as a `ModelIdentifierField`.
-
-        ui_model_variant: Specifies the model variant(s) to filter the model list by in the Workflow Editor. For example,
-        `ui_model_variant=ModelVariantType.Inpainting` will show only inpainting models. This arg is only valid if this
-        Input field is annotated as a `ModelIdentifierField`.
-
-        ui_model_format: Specifies the model format(s) to filter the model list by in the Workflow Editor. For example,
-        `ui_model_format=ModelFormat.Diffusers` will show only models in the diffusers format. This arg is only valid
-        if this Input field is annotated as a `ModelIdentifierField`.
-
-        ui_choice_labels: Specifies the labels to use for the choices in an enum field. If omitted, the enum values
-        will be used. This arg is only valid if the field is annotated with as a `Literal`. For example,
-        `Literal["choice1", "choice2", "choice3"]` with `ui_choice_labels={"choice1": "Choice 1", "choice2": "Choice 2",
-        "choice3": "Choice 3"}` will render a dropdown with the labels "Choice 1", "Choice 2" and "Choice 3".
+    :param dict[str, str] ui_choice_labels: [None] Specifies the labels to use for the choices in an enum field.
     """
 
     json_schema_extra_ = InputFieldJSONSchemaExtra(
@@ -582,92 +538,7 @@ def InputField(
     )
 
     if ui_type is not None:
-        if (
-            ui_model_base is not None
-            or ui_model_type is not None
-            or ui_model_variant is not None
-            or ui_model_format is not None
-        ):
-            logger.warning("InputField: Use either ui_type or ui_model_[base|type|variant|format]. Ignoring ui_type.")
-        # Map old-style UIType to new-style ui_model_[base|type|variant|format]
-        elif ui_type is UIType.MainModel:
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.CogView4MainModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.CogView4]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.FluxMainModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.Flux]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.SD3MainModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.StableDiffusion3]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.SDXLMainModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.StableDiffusionXL]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.SDXLRefinerModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.StableDiffusionXLRefiner]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        # Think this UIType is unused...?
-        # elif ui_type is UIType.ONNXModel:
-        #     json_schema_extra_.ui_model_base =
-        #     json_schema_extra_.ui_model_type =
-        elif ui_type is UIType.VAEModel:
-            json_schema_extra_.ui_model_type = [ModelType.VAE]
-        elif ui_type is UIType.FluxVAEModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.Flux]
-            json_schema_extra_.ui_model_type = [ModelType.VAE]
-        elif ui_type is UIType.LoRAModel:
-            json_schema_extra_.ui_model_type = [ModelType.LoRA]
-        elif ui_type is UIType.ControlNetModel:
-            json_schema_extra_.ui_model_type = [ModelType.ControlNet]
-        elif ui_type is UIType.IPAdapterModel:
-            json_schema_extra_.ui_model_type = [ModelType.IPAdapter]
-        elif ui_type is UIType.T2IAdapterModel:
-            json_schema_extra_.ui_model_type = [ModelType.T2IAdapter]
-        elif ui_type is UIType.T5EncoderModel:
-            json_schema_extra_.ui_model_type = [ModelType.T5Encoder]
-        elif ui_type is UIType.CLIPEmbedModel:
-            json_schema_extra_.ui_model_type = [ModelType.CLIPEmbed]
-        elif ui_type is UIType.CLIPLEmbedModel:
-            json_schema_extra_.ui_model_type = [ModelType.CLIPEmbed]
-            json_schema_extra_.ui_model_variant = [ClipVariantType.L]
-        elif ui_type is UIType.CLIPGEmbedModel:
-            json_schema_extra_.ui_model_type = [ModelType.CLIPEmbed]
-            json_schema_extra_.ui_model_variant = [ClipVariantType.G]
-        elif ui_type is UIType.SpandrelImageToImageModel:
-            json_schema_extra_.ui_model_type = [ModelType.SpandrelImageToImage]
-        elif ui_type is UIType.ControlLoRAModel:
-            json_schema_extra_.ui_model_type = [ModelType.ControlLoRa]
-        elif ui_type is UIType.SigLipModel:
-            json_schema_extra_.ui_model_type = [ModelType.SigLIP]
-        elif ui_type is UIType.FluxReduxModel:
-            json_schema_extra_.ui_model_type = [ModelType.FluxRedux]
-        elif ui_type is UIType.LlavaOnevisionModel:
-            json_schema_extra_.ui_model_type = [ModelType.LlavaOnevision]
-        elif ui_type is UIType.Imagen3Model:
-            json_schema_extra_.ui_model_base = [BaseModelType.Imagen3]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.Imagen4Model:
-            json_schema_extra_.ui_model_base = [BaseModelType.Imagen4]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.ChatGPT4oModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.ChatGPT4o]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.Gemini2_5Model:
-            json_schema_extra_.ui_model_base = [BaseModelType.Gemini2_5]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.FluxKontextModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.FluxKontext]
-            json_schema_extra_.ui_model_type = [ModelType.Main]
-        elif ui_type is UIType.Veo3Model:
-            json_schema_extra_.ui_model_base = [BaseModelType.Veo3]
-            json_schema_extra_.ui_model_type = [ModelType.Video]
-        elif ui_type is UIType.RunwayModel:
-            json_schema_extra_.ui_model_base = [BaseModelType.Runway]
-            json_schema_extra_.ui_model_type = [ModelType.Video]
-        else:
-            json_schema_extra_.ui_type = ui_type
-
+        json_schema_extra_.ui_type = ui_type
     if ui_component is not None:
         json_schema_extra_.ui_component = ui_component
     if ui_hidden is not None:
@@ -676,26 +547,6 @@ def InputField(
         json_schema_extra_.ui_order = ui_order
     if ui_choice_labels is not None:
         json_schema_extra_.ui_choice_labels = ui_choice_labels
-    if ui_model_base is not None:
-        if isinstance(ui_model_base, list):
-            json_schema_extra_.ui_model_base = ui_model_base
-        else:
-            json_schema_extra_.ui_model_base = [ui_model_base]
-    if ui_model_type is not None:
-        if isinstance(ui_model_type, list):
-            json_schema_extra_.ui_model_type = ui_model_type
-        else:
-            json_schema_extra_.ui_model_type = [ui_model_type]
-    if ui_model_variant is not None:
-        if isinstance(ui_model_variant, list):
-            json_schema_extra_.ui_model_variant = ui_model_variant
-        else:
-            json_schema_extra_.ui_model_variant = [ui_model_variant]
-    if ui_model_format is not None:
-        if isinstance(ui_model_format, list):
-            json_schema_extra_.ui_model_format = ui_model_format
-        else:
-            json_schema_extra_.ui_model_format = [ui_model_format]
 
     """
     There is a conflict between the typing of invocation definitions and the typing of an invocation's
@@ -797,20 +648,20 @@ def OutputField(
     """
     Creates an output field for an invocation output.
 
-    This is a wrapper for Pydantic's [Field](https://docs.pydantic.dev/1.10/usage/schema/#field-customization)
+    This is a wrapper for Pydantic's [Field](https://docs.pydantic.dev/1.10/usage/schema/#field-customization) \
     that adds a few extra parameters to support graph execution and the node editor UI.
 
-    Args:
-        ui_type: Optionally provides an extra type hint for the UI. In some situations, the field's type is not enough
-        to infer the correct UI type. For example, Scheduler fields are enums, but we want to render a special scheduler
-        dropdown in the UI. Use `UIType.Scheduler` to indicate this.
+    :param UIType ui_type: [None] Optionally provides an extra type hint for the UI. \
+      In some situations, the field's type is not enough to infer the correct UI type. \
+      For example, model selection fields should render a dropdown UI component to select a model. \
+      Internally, there is no difference between SD-1, SD-2 and SDXL model fields, they all use \
+      `MainModelField`. So to ensure the base-model-specific UI is rendered, you can use \
+      `UIType.SDXLMainModelField` to indicate that the field is an SDXL main model field.
 
-        ui_hidden: Specifies whether or not this field should be hidden in the UI.
+    :param bool ui_hidden: [False] Specifies whether or not this field should be hidden in the UI. \
 
-        ui_order: Specifies the order in which this field should be rendered in the UI. If omitted, the field will be
-        rendered after all fields with an explicit order, in the order they are defined in the Invocation class.
+    :param int ui_order: [None] Specifies the order in which this field should be rendered in the UI. \
     """
-
     return Field(
         default=default,
         title=title,
@@ -828,9 +679,9 @@ def OutputField(
         min_length=min_length,
         max_length=max_length,
         json_schema_extra=OutputFieldJSONSchemaExtra(
+            ui_type=ui_type,
             ui_hidden=ui_hidden,
             ui_order=ui_order,
-            ui_type=ui_type,
             field_kind=FieldKind.Output,
         ).model_dump(exclude_none=True),
     )

invokeai/app/invocations/flux_control_lora_loader.py

@@ -4,10 +4,9 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, OutputField, UIType
 from invokeai.app.invocations.model import ControlLoRAField, ModelIdentifierField
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType
 
 
 @invocation_output("flux_control_lora_loader_output")
@@ -30,10 +29,7 @@ class FluxControlLoRALoaderInvocation(BaseInvocation):
     """LoRA model and Image to use with FLUX transformer generation."""
 
     lora: ModelIdentifierField = InputField(
-        description=FieldDescriptions.control_lora_model,
-        title="Control LoRA",
-        ui_model_base=BaseModelType.Flux,
-        ui_model_type=ModelType.ControlLoRa,
+        description=FieldDescriptions.control_lora_model, title="Control LoRA", ui_type=UIType.ControlLoRAModel
     )
     image: ImageField = InputField(description="The image to encode.")
     weight: float = InputField(description="The weight of the LoRA.", default=1.0)

invokeai/app/invocations/flux_controlnet.py

@@ -6,12 +6,11 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, OutputField, UIType
 from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.util import validate_begin_end_step, validate_weights
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.controlnet_utils import CONTROLNET_RESIZE_VALUES
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType
 
 
 class FluxControlNetField(BaseModel):
@@ -58,9 +57,7 @@ class FluxControlNetInvocation(BaseInvocation):
 
     image: ImageField = InputField(description="The control image")
     control_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.controlnet_model,
-        ui_model_base=BaseModelType.Flux,
-        ui_model_type=ModelType.ControlNet,
+        description=FieldDescriptions.controlnet_model, ui_type=UIType.ControlNetModel
     )
     control_weight: float | list[float] = InputField(
         default=1.0, ge=-1, le=2, description="The weight given to the ControlNet"

invokeai/app/invocations/flux_ip_adapter.py

@@ -5,7 +5,7 @@ from pydantic import field_validator, model_validator
 from typing_extensions import Self
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
-from invokeai.app.invocations.fields import InputField
+from invokeai.app.invocations.fields import InputField, UIType
 from invokeai.app.invocations.ip_adapter import (
     CLIP_VISION_MODEL_MAP,
     IPAdapterField,
@@ -20,7 +20,6 @@ from invokeai.backend.model_manager.config import (
     IPAdapterCheckpointConfig,
     IPAdapterInvokeAIConfig,
 )
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType
 
 
 @invocation(
@@ -37,10 +36,7 @@ class FluxIPAdapterInvocation(BaseInvocation):
 
     image: ImageField = InputField(description="The IP-Adapter image prompt(s).")
     ip_adapter_model: ModelIdentifierField = InputField(
-        description="The IP-Adapter model.",
-        title="IP-Adapter Model",
-        ui_model_base=BaseModelType.Flux,
-        ui_model_type=ModelType.IPAdapter,
+        description="The IP-Adapter model.", title="IP-Adapter Model", ui_type=UIType.IPAdapterModel
     )
     # Currently, the only known ViT model used by FLUX IP-Adapters is ViT-L.
     clip_vision_model: Literal["ViT-L"] = InputField(description="CLIP Vision model to use.", default="ViT-L")

invokeai/app/invocations/flux_lora_loader.py

@@ -6,10 +6,10 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
 from invokeai.app.invocations.model import CLIPField, LoRAField, ModelIdentifierField, T5EncoderField, TransformerField
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType
+from invokeai.backend.model_manager.taxonomy import BaseModelType
 
 
 @invocation_output("flux_lora_loader_output")
@@ -36,10 +36,7 @@ class FluxLoRALoaderInvocation(BaseInvocation):
     """Apply a LoRA model to a FLUX transformer and/or text encoder."""
 
     lora: ModelIdentifierField = InputField(
-        description=FieldDescriptions.lora_model,
-        title="LoRA",
-        ui_model_base=BaseModelType.Flux,
-        ui_model_type=ModelType.LoRA,
+        description=FieldDescriptions.lora_model, title="LoRA", ui_type=UIType.LoRAModel
     )
     weight: float = InputField(default=0.75, description=FieldDescriptions.lora_weight)
     transformer: TransformerField | None = InputField(

invokeai/app/invocations/flux_model_loader.py

@@ -6,7 +6,7 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
 from invokeai.app.invocations.model import CLIPField, ModelIdentifierField, T5EncoderField, TransformerField, VAEField
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.t5_model_identifier import (
@@ -17,7 +17,7 @@ from invokeai.backend.flux.util import max_seq_lengths
 from invokeai.backend.model_manager.config import (
     CheckpointConfigBase,
 )
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType, SubModelType
+from invokeai.backend.model_manager.taxonomy import SubModelType
 
 
 @invocation_output("flux_model_loader_output")
@@ -46,30 +46,23 @@ class FluxModelLoaderInvocation(BaseInvocation):
 
     model: ModelIdentifierField = InputField(
         description=FieldDescriptions.flux_model,
+        ui_type=UIType.FluxMainModel,
         input=Input.Direct,
-        ui_model_base=BaseModelType.Flux,
-        ui_model_type=ModelType.Main,
     )
 
     t5_encoder_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.t5_encoder,
-        input=Input.Direct,
-        title="T5 Encoder",
-        ui_model_type=ModelType.T5Encoder,
+        description=FieldDescriptions.t5_encoder, ui_type=UIType.T5EncoderModel, input=Input.Direct, title="T5 Encoder"
     )
 
     clip_embed_model: ModelIdentifierField = InputField(
         description=FieldDescriptions.clip_embed_model,
+        ui_type=UIType.CLIPEmbedModel,
         input=Input.Direct,
         title="CLIP Embed",
-        ui_model_type=ModelType.CLIPEmbed,
     )
 
     vae_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.vae_model,
-        title="VAE",
-        ui_model_base=BaseModelType.Flux,
-        ui_model_type=ModelType.VAE,
+        description=FieldDescriptions.vae_model, ui_type=UIType.FluxVAEModel, title="VAE"
     )
 
     def invoke(self, context: InvocationContext) -> FluxModelLoaderOutput:

invokeai/app/invocations/flux_redux.py

@@ -18,6 +18,7 @@ from invokeai.app.invocations.fields import (
     InputField,
     OutputField,
     TensorField,
+    UIType,
 )
 from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.primitives import ImageField
@@ -63,8 +64,7 @@ class FluxReduxInvocation(BaseInvocation):
     redux_model: ModelIdentifierField = InputField(
         description="The FLUX Redux model to use.",
         title="FLUX Redux Model",
-        ui_model_base=BaseModelType.Flux,
-        ui_model_type=ModelType.FluxRedux,
+        ui_type=UIType.FluxReduxModel,
     )
     downsampling_factor: int = InputField(
         ge=1,

invokeai/app/invocations/ip_adapter.py

@@ -5,7 +5,7 @@ from pydantic import BaseModel, Field, field_validator, model_validator
 from typing_extensions import Self
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, BaseInvocationOutput, invocation, invocation_output
-from invokeai.app.invocations.fields import FieldDescriptions, InputField, OutputField, TensorField
+from invokeai.app.invocations.fields import FieldDescriptions, InputField, OutputField, TensorField, UIType
 from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.primitives import ImageField
 from invokeai.app.invocations.util import validate_begin_end_step, validate_weights
@@ -85,8 +85,7 @@ class IPAdapterInvocation(BaseInvocation):
         description="The IP-Adapter model.",
         title="IP-Adapter Model",
         ui_order=-1,
-        ui_model_base=[BaseModelType.StableDiffusion1, BaseModelType.StableDiffusionXL],
-        ui_model_type=ModelType.IPAdapter,
+        ui_type=UIType.IPAdapterModel,
     )
     clip_vision_model: Literal["ViT-H", "ViT-G", "ViT-L"] = InputField(
         description="CLIP Vision model to use. Overrides model settings. Mandatory for checkpoint models.",

invokeai/app/invocations/llava_onevision_vllm.py

@@ -6,12 +6,11 @@ from pydantic import field_validator
 from transformers import AutoProcessor, LlavaOnevisionForConditionalGeneration, LlavaOnevisionProcessor
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
-from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, UIComponent
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, UIComponent, UIType
 from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.primitives import StringOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.backend.llava_onevision_pipeline import LlavaOnevisionPipeline
-from invokeai.backend.model_manager.taxonomy import ModelType
 from invokeai.backend.util.devices import TorchDevice
 
 
@@ -35,7 +34,7 @@ class LlavaOnevisionVllmInvocation(BaseInvocation):
     vllm_model: ModelIdentifierField = InputField(
         title="LLaVA Model Type",
         description=FieldDescriptions.vllm_model,
-        ui_model_type=ModelType.LlavaOnevision,
+        ui_type=UIType.LlavaOnevisionModel,
     )
 
     @field_validator("images", mode="before")

invokeai/app/invocations/metadata_linked.py

@@ -53,7 +53,7 @@ from invokeai.app.invocations.primitives import (
 from invokeai.app.invocations.scheduler import SchedulerOutput
 from invokeai.app.invocations.t2i_adapter import T2IAdapterField, T2IAdapterInvocation
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType, SubModelType
+from invokeai.backend.model_manager.taxonomy import ModelType, SubModelType
 from invokeai.backend.stable_diffusion.schedulers.schedulers import SCHEDULER_NAME_VALUES
 from invokeai.version import __version__
 
@@ -473,6 +473,7 @@ class MetadataToModelOutput(BaseInvocationOutput):
     model: ModelIdentifierField = OutputField(
         description=FieldDescriptions.main_model,
         title="Model",
+        ui_type=UIType.MainModel,
     )
     name: str = OutputField(description="Model Name", title="Name")
     unet: UNetField = OutputField(description=FieldDescriptions.unet, title="UNet")
@@ -487,6 +488,7 @@ class MetadataToSDXLModelOutput(BaseInvocationOutput):
     model: ModelIdentifierField = OutputField(
         description=FieldDescriptions.main_model,
         title="Model",
+        ui_type=UIType.SDXLMainModel,
     )
     name: str = OutputField(description="Model Name", title="Name")
     unet: UNetField = OutputField(description=FieldDescriptions.unet, title="UNet")
@@ -517,7 +519,8 @@ class MetadataToModelInvocation(BaseInvocation, WithMetadata):
         input=Input.Direct,
     )
     default_value: ModelIdentifierField = InputField(
-        description="The default model to use if not found in the metadata", ui_model_type=ModelType.Main
+        description="The default model to use if not found in the metadata",
+        ui_type=UIType.MainModel,
     )
 
     _validate_custom_label = model_validator(mode="after")(validate_custom_label)
@@ -572,8 +575,7 @@ class MetadataToSDXLModelInvocation(BaseInvocation, WithMetadata):
     )
     default_value: ModelIdentifierField = InputField(
         description="The default SDXL Model to use if not found in the metadata",
-        ui_model_type=ModelType.Main,
-        ui_model_base=BaseModelType.StableDiffusionXL,
+        ui_type=UIType.SDXLMainModel,
     )
 
     _validate_custom_label = model_validator(mode="after")(validate_custom_label)

invokeai/app/invocations/model.py

@@ -9,7 +9,7 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, ImageField, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, Input, InputField, OutputField, UIType
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.shared.models import FreeUConfig
 from invokeai.backend.model_manager.config import (
@@ -73,12 +73,6 @@ class GlmEncoderField(BaseModel):
     text_encoder: ModelIdentifierField = Field(description="Info to load text_encoder submodel")
 
 
-class Qwen2_5VLField(BaseModel):
-    tokenizer: ModelIdentifierField = Field(description="Info to load Qwen2.5-VL tokenizer submodel")
-    text_encoder: ModelIdentifierField = Field(description="Info to load Qwen2.5-VL text encoder submodel")
-    loras: List[LoRAField] = Field(default_factory=list, description="LoRAs to apply on model loading")
-
-
 class VAEField(BaseModel):
     vae: ModelIdentifierField = Field(description="Info to load vae submodel")
     seamless_axes: List[str] = Field(default_factory=list, description='Axes("x" and "y") to which apply seamless')
@@ -151,7 +145,7 @@ class ModelIdentifierInvocation(BaseInvocation):
 
 @invocation(
     "main_model_loader",
-    title="Main Model - SD1.5, SD2",
+    title="Main Model - SD1.5",
     tags=["model"],
     category="model",
     version="1.0.4",
@@ -159,11 +153,7 @@ class ModelIdentifierInvocation(BaseInvocation):
 class MainModelLoaderInvocation(BaseInvocation):
     """Loads a main model, outputting its submodels."""
 
-    model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.main_model,
-        ui_model_base=[BaseModelType.StableDiffusion1, BaseModelType.StableDiffusion2],
-        ui_model_type=ModelType.Main,
-    )
+    model: ModelIdentifierField = InputField(description=FieldDescriptions.main_model, ui_type=UIType.MainModel)
     # TODO: precision?
 
     def invoke(self, context: InvocationContext) -> ModelLoaderOutput:
@@ -197,10 +187,7 @@ class LoRALoaderInvocation(BaseInvocation):
     """Apply selected lora to unet and text_encoder."""
 
     lora: ModelIdentifierField = InputField(
-        description=FieldDescriptions.lora_model,
-        title="LoRA",
-        ui_model_base=BaseModelType.StableDiffusion1,
-        ui_model_type=ModelType.LoRA,
+        description=FieldDescriptions.lora_model, title="LoRA", ui_type=UIType.LoRAModel
     )
     weight: float = InputField(default=0.75, description=FieldDescriptions.lora_weight)
     unet: Optional[UNetField] = InputField(
@@ -263,9 +250,7 @@ class LoRASelectorInvocation(BaseInvocation):
     """Selects a LoRA model and weight."""
 
     lora: ModelIdentifierField = InputField(
-        description=FieldDescriptions.lora_model,
-        title="LoRA",
-        ui_model_type=ModelType.LoRA,
+        description=FieldDescriptions.lora_model, title="LoRA", ui_type=UIType.LoRAModel
     )
     weight: float = InputField(default=0.75, description=FieldDescriptions.lora_weight)
 
@@ -347,10 +332,7 @@ class SDXLLoRALoaderInvocation(BaseInvocation):
     """Apply selected lora to unet and text_encoder."""
 
     lora: ModelIdentifierField = InputField(
-        description=FieldDescriptions.lora_model,
-        title="LoRA",
-        ui_model_base=BaseModelType.StableDiffusionXL,
-        ui_model_type=ModelType.LoRA,
+        description=FieldDescriptions.lora_model, title="LoRA", ui_type=UIType.LoRAModel
     )
     weight: float = InputField(default=0.75, description=FieldDescriptions.lora_weight)
     unet: Optional[UNetField] = InputField(
@@ -491,26 +473,13 @@ class SDXLLoRACollectionLoader(BaseInvocation):
 
 
 @invocation(
-    "vae_loader",
-    title="VAE Model - SD1.5, SD2, SDXL, SD3, FLUX",
-    tags=["vae", "model"],
-    category="model",
-    version="1.0.4",
+    "vae_loader", title="VAE Model - SD1.5, SDXL, SD3, FLUX", tags=["vae", "model"], category="model", version="1.0.4"
 )
 class VAELoaderInvocation(BaseInvocation):
     """Loads a VAE model, outputting a VaeLoaderOutput"""
 
     vae_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.vae_model,
-        title="VAE",
-        ui_model_base=[
-            BaseModelType.StableDiffusion1,
-            BaseModelType.StableDiffusion2,
-            BaseModelType.StableDiffusionXL,
-            BaseModelType.StableDiffusion3,
-            BaseModelType.Flux,
-        ],
-        ui_model_type=ModelType.VAE,
+        description=FieldDescriptions.vae_model, title="VAE", ui_type=UIType.VAEModel
     )
 
     def invoke(self, context: InvocationContext) -> VAEOutput:

invokeai/app/invocations/primitives.py

@@ -24,7 +24,6 @@ from invokeai.app.invocations.fields import (
     InputField,
     LatentsField,
     OutputField,
-    QwenImageConditioningField,
     SD3ConditioningField,
     TensorField,
     UIComponent,
@@ -487,17 +486,6 @@ class CogView4ConditioningOutput(BaseInvocationOutput):
         return cls(conditioning=CogView4ConditioningField(conditioning_name=conditioning_name))
 
 
-@invocation_output("qwen_image_conditioning_output")
-class QwenImageConditioningOutput(BaseInvocationOutput):
-    """Base class for nodes that output a Qwen-Image conditioning tensor."""
-
-    conditioning: QwenImageConditioningField = OutputField(description=FieldDescriptions.cond)
-
-    @classmethod
-    def build(cls, conditioning_name: str) -> "QwenImageConditioningOutput":
-        return cls(conditioning=QwenImageConditioningField(conditioning_name=conditioning_name))
-
-
 @invocation_output("conditioning_output")
 class ConditioningOutput(BaseInvocationOutput):
     """Base class for nodes that output a single conditioning tensor"""

invokeai/app/invocations/qwen_image_denoise.py

@@ -1,150 +0,0 @@
-# Copyright (c) 2024, Brandon W. Rising and the InvokeAI Development Team
-"""Qwen-Image denoising invocation using diffusers pipeline."""
-
-
-import torch
-
-from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
-from invokeai.app.invocations.fields import (
-    FieldDescriptions,
-    Input,
-    InputField,
-    QwenImageConditioningField,
-    WithBoard,
-    WithMetadata,
-)
-from invokeai.app.invocations.model import TransformerField, VAEField
-from invokeai.app.invocations.primitives import ImageOutput
-from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.util.devices import TorchDevice
-
-
-@invocation(
-    "qwen_image_denoise",
-    title="Qwen-Image Denoise",
-    tags=["image", "qwen"],
-    category="image",
-    version="1.0.0",
-)
-class QwenImageDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
-    """Run text-to-image generation with a Qwen-Image diffusion model."""
-
-    # Model components
-    transformer: TransformerField = InputField(
-        description=FieldDescriptions.transformer,
-        input=Input.Connection,
-        title="Transformer",
-    )
-
-    vae: VAEField = InputField(
-        description=FieldDescriptions.vae,
-        input=Input.Connection,
-        title="VAE",
-    )
-
-    # Text conditioning
-    positive_conditioning: QwenImageConditioningField = InputField(
-        description=FieldDescriptions.positive_cond, input=Input.Connection
-    )
-
-    # Generation parameters
-    width: int = InputField(default=1024, multiple_of=16, description="Width of the generated image.")
-    height: int = InputField(default=1024, multiple_of=16, description="Height of the generated image.")
-    num_inference_steps: int = InputField(
-        default=50, gt=0, description="Number of denoising steps."
-    )
-    guidance_scale: float = InputField(
-        default=7.5, gt=1.0, description="Classifier-free guidance scale."
-    )
-    seed: int = InputField(default=0, description="Randomness seed for reproducibility.")
-
-    @torch.no_grad()
-    def invoke(self, context: InvocationContext) -> ImageOutput:
-        """Generate image using Qwen-Image pipeline."""
-
-        device = TorchDevice.choose_torch_device()
-        dtype = torch.bfloat16 if torch.cuda.is_available() else torch.float32
-
-        # Load model components
-        with context.models.load(self.transformer.transformer) as transformer_info, \
-             context.models.load(self.vae.vae) as vae_info:
-
-            # Load conditioning data
-            conditioning_data = context.conditioning.load(self.positive_conditioning.conditioning_name)
-            assert len(conditioning_data.conditionings) == 1
-            conditioning_info = conditioning_data.conditionings[0]
-
-            # Extract the prompt from conditioning
-            # The text encoder node stores both embeddings and the original prompt
-            prompt = getattr(conditioning_info, 'prompt', "A high-quality image")
-
-            # For now, we'll create a simplified pipeline
-            # In a full implementation, we'd properly load all components
-            try:
-                # Create the Qwen-Image pipeline with loaded components
-                # Note: This is a simplified approach. In production, we'd need to:
-                # 1. Load the text encoder from the conditioning
-                # 2. Properly initialize the pipeline with all components
-                # 3. Handle model configuration and dtype conversion
-
-                # For demonstration, we'll assume the models are loaded correctly
-                # and create a basic generation
-                transformer_model = transformer_info.model
-                vae_model = vae_info.model
-
-                # Move models to device
-                transformer_model = transformer_model.to(device, dtype=dtype)
-                vae_model = vae_model.to(device, dtype=dtype)
-
-                # Set up generator for reproducibility
-                generator = torch.Generator(device=device)
-                generator.manual_seed(self.seed)
-
-                # Create latents
-                latent_shape = (
-                    1,
-                    vae_model.config.latent_channels if hasattr(vae_model, 'config') else 4,
-                    self.height // 8,
-                    self.width // 8,
-                )
-                latents = torch.randn(latent_shape, generator=generator, device=device, dtype=dtype)
-
-                # Simple denoising loop (placeholder for actual implementation)
-                # In reality, we'd use the full QwenImagePipeline or implement the proper denoising
-                for _ in range(self.num_inference_steps):
-                    # This is a placeholder - actual implementation would:
-                    # 1. Apply noise scheduling
-                    # 2. Use the transformer for denoising
-                    # 3. Apply guidance scale
-                    latents = latents * 0.99  # Placeholder denoising
-
-                # Decode latents to image
-                with torch.no_grad():
-                    # Scale latents
-                    latents = latents / vae_model.config.scaling_factor if hasattr(vae_model, 'config') else latents
-                    # Decode
-                    image = vae_model.decode(latents).sample if hasattr(vae_model, 'decode') else latents
-
-                    # Convert to PIL Image
-                    image = (image / 2 + 0.5).clamp(0, 1)
-                    image = image.cpu().permute(0, 2, 3, 1).float().numpy()
-
-                    if image.ndim == 4:
-                        image = image[0]
-
-                    # Convert to uint8
-                    image = (image * 255).round().astype("uint8")
-
-                    # Convert numpy array to PIL Image
-                    from PIL import Image
-                    pil_image = Image.fromarray(image)
-
-            except Exception as e:
-                context.logger.error(f"Error during Qwen-Image generation: {e}")
-                # Create a placeholder image on error
-                from PIL import Image
-                pil_image = Image.new('RGB', (self.width, self.height), color='gray')
-
-            # Save and return the generated image
-            image_dto = context.images.save(image=pil_image)
-            return ImageOutput.build(image_dto)

invokeai/app/invocations/qwen_image_model_loader.py

@@ -1,83 +0,0 @@
-from invokeai.app.invocations.baseinvocation import (
-    BaseInvocation,
-    BaseInvocationOutput,
-    invocation,
-    invocation_output,
-)
-from invokeai.app.invocations.fields import Input, InputField, OutputField
-from invokeai.app.invocations.model import ModelIdentifierField, Qwen2_5VLField, TransformerField, VAEField
-from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType, SubModelType
-
-
-@invocation_output("qwen_image_model_loader_output")
-class QwenImageModelLoaderOutput(BaseInvocationOutput):
-    """Qwen-Image base model loader output"""
-
-    transformer: TransformerField = OutputField(description="Qwen-Image transformer model", title="Transformer")
-    qwen2_5_vl: Qwen2_5VLField = OutputField(description="Qwen2.5-VL text encoder for Qwen-Image", title="Text Encoder")
-    vae: VAEField = OutputField(description="Qwen-Image VAE", title="VAE")
-
-
-@invocation(
-    "qwen_image_model_loader",
-    title="Main Model - Qwen-Image",
-    tags=["model", "qwen-image"],
-    category="model",
-    version="1.0.0",
-)
-class QwenImageModelLoaderInvocation(BaseInvocation):
-    """Loads a Qwen-Image base model, outputting its submodels."""
-
-    model: ModelIdentifierField = InputField(
-        description="Qwen-Image main model",
-        input=Input.Direct,
-        ui_model_base=BaseModelType.QwenImage,
-        ui_model_type=ModelType.Main,
-    )
-
-    qwen2_5_vl_model: ModelIdentifierField = InputField(
-        description="Qwen2.5-VL vision-language model",
-        input=Input.Direct,
-        title="Qwen2.5-VL Model",
-        ui_model_base=BaseModelType.QwenImage,
-        # ui_model_type=ModelType.VL
-    )
-
-    vae_model: ModelIdentifierField | None = InputField(
-        description="VAE model for Qwen-Image",
-        title="VAE",
-        ui_model_base=BaseModelType.QwenImage,
-        ui_model_type=ModelType.VAE,
-        default=None,
-    )
-
-    def invoke(self, context: InvocationContext) -> QwenImageModelLoaderOutput:
-        # Validate that required models exist
-        for key in [self.model.key, self.qwen2_5_vl_model.key]:
-            if not context.models.exists(key):
-                raise ValueError(f"Unknown model: {key}")
-
-        # Validate optional VAE model if provided
-        if self.vae_model and not context.models.exists(self.vae_model.key):
-            raise ValueError(f"Unknown model: {self.vae_model.key}")
-
-        # Create submodel references
-        transformer = self.model.model_copy(update={"submodel_type": SubModelType.Transformer})
-
-        # Use provided VAE or extract from main model
-        if self.vae_model:
-            vae = self.vae_model.model_copy(update={"submodel_type": SubModelType.VAE})
-        else:
-            # Use the VAE bundled with the Qwen-Image model
-            vae = self.model.model_copy(update={"submodel_type": SubModelType.VAE})
-
-        # For Qwen-Image, we use Qwen2.5-VL as the text encoder
-        tokenizer = self.qwen2_5_vl_model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
-        text_encoder = self.qwen2_5_vl_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
-
-        return QwenImageModelLoaderOutput(
-            transformer=TransformerField(transformer=transformer, loras=[]),
-            qwen2_5_vl=Qwen2_5VLField(tokenizer=tokenizer, text_encoder=text_encoder, loras=[]),
-            vae=VAEField(vae=vae),
-        )

invokeai/app/invocations/qwen_image_text_encoder.py

@@ -1,79 +0,0 @@
-# Copyright (c) 2024, Brandon W. Rising and the InvokeAI Development Team
-"""Qwen-Image text encoding invocation."""
-
-import torch
-
-from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
-from invokeai.app.invocations.fields import Input, InputField, UIComponent
-from invokeai.app.invocations.model import Qwen2_5VLField
-from invokeai.app.invocations.primitives import QwenImageConditioningOutput
-from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.stable_diffusion.diffusion.conditioning_data import ConditioningFieldData
-
-
-@invocation(
-    "qwen_image_text_encoder",
-    title="Prompt - Qwen-Image",
-    tags=["prompt", "conditioning", "qwen"],
-    category="conditioning",
-    version="1.0.0",
-)
-class QwenImageTextEncoderInvocation(BaseInvocation):
-    """Encodes a text prompt for Qwen-Image generation."""
-
-    prompt: str = InputField(description="Text prompt to encode.", ui_component=UIComponent.Textarea)
-    qwen2_5_vl: Qwen2_5VLField = InputField(
-        title="Qwen2.5-VL",
-        description="Qwen2.5-VL vision-language model for text encoding",
-        input=Input.Connection,
-    )
-
-    @torch.no_grad()
-    def invoke(self, context: InvocationContext) -> QwenImageConditioningOutput:
-        """Encode the prompt using Qwen-Image's text encoder."""
-
-        # Load the text encoder info first to get the model
-        text_encoder_info = context.models.load(self.qwen2_5_vl.text_encoder)
-
-        # Load the Qwen2.5-VL tokenizer and text encoder with proper device management
-        with text_encoder_info.model_on_device() as (cached_weights, text_encoder), \
-             context.models.load(self.qwen2_5_vl.tokenizer) as tokenizer:
-
-            try:
-                # Tokenize the prompt
-                # Qwen2.5-VL supports much longer sequences than CLIP
-                text_inputs = tokenizer(
-                    self.prompt,
-                    padding="max_length",
-                    max_length=1024,  # Qwen2.5-VL supports much longer sequences
-                    truncation=True,
-                    return_tensors="pt",
-                )
-
-                # Encode the text (text_encoder is already on the correct device)
-                text_embeddings = text_encoder(text_inputs.input_ids.to(text_encoder.device))[0]
-
-                # Create a simple conditioning info that stores the embeddings
-                # For now, we'll create a simple class to hold the data
-                class QwenImageConditioningInfo:
-                    def __init__(self, text_embeds: torch.Tensor, prompt: str):
-                        self.text_embeds = text_embeds
-                        self.prompt = prompt
-
-                conditioning_info = QwenImageConditioningInfo(text_embeddings, self.prompt)
-                conditioning_data = ConditioningFieldData(conditionings=[conditioning_info])
-
-                conditioning_name = context.conditioning.save(conditioning_data)
-                return QwenImageConditioningOutput.build(conditioning_name)
-
-            except Exception as e:
-                context.logger.error(f"Error encoding Qwen-Image text: {e}")
-                # Fallback to simple text storage
-                class QwenImageConditioningInfo:
-                    def __init__(self, prompt: str):
-                        self.prompt = prompt
-
-                conditioning_info = QwenImageConditioningInfo(self.prompt)
-                conditioning_data = ConditioningFieldData(conditionings=[conditioning_info])
-                conditioning_name = context.conditioning.save(conditioning_data)
-                return QwenImageConditioningOutput.build(conditioning_name)

invokeai/app/invocations/sd3_model_loader.py

@@ -6,14 +6,14 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
 from invokeai.app.invocations.model import CLIPField, ModelIdentifierField, T5EncoderField, TransformerField, VAEField
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.t5_model_identifier import (
     preprocess_t5_encoder_model_identifier,
     preprocess_t5_tokenizer_model_identifier,
 )
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ClipVariantType, ModelType, SubModelType
+from invokeai.backend.model_manager.taxonomy import SubModelType
 
 
 @invocation_output("sd3_model_loader_output")
@@ -39,43 +39,36 @@ class Sd3ModelLoaderInvocation(BaseInvocation):
 
     model: ModelIdentifierField = InputField(
         description=FieldDescriptions.sd3_model,
+        ui_type=UIType.SD3MainModel,
         input=Input.Direct,
-        ui_model_base=BaseModelType.StableDiffusion3,
-        ui_model_type=ModelType.Main,
     )
 
     t5_encoder_model: Optional[ModelIdentifierField] = InputField(
         description=FieldDescriptions.t5_encoder,
+        ui_type=UIType.T5EncoderModel,
         input=Input.Direct,
         title="T5 Encoder",
         default=None,
-        ui_model_type=ModelType.T5Encoder,
     )
 
     clip_l_model: Optional[ModelIdentifierField] = InputField(
         description=FieldDescriptions.clip_embed_model,
+        ui_type=UIType.CLIPLEmbedModel,
         input=Input.Direct,
         title="CLIP L Encoder",
         default=None,
-        ui_model_type=ModelType.CLIPEmbed,
-        ui_model_variant=ClipVariantType.L,
     )
 
     clip_g_model: Optional[ModelIdentifierField] = InputField(
         description=FieldDescriptions.clip_g_model,
+        ui_type=UIType.CLIPGEmbedModel,
         input=Input.Direct,
         title="CLIP G Encoder",
         default=None,
-        ui_model_type=ModelType.CLIPEmbed,
-        ui_model_variant=ClipVariantType.G,
     )
 
     vae_model: Optional[ModelIdentifierField] = InputField(
-        description=FieldDescriptions.vae_model,
-        title="VAE",
-        default=None,
-        ui_model_base=BaseModelType.StableDiffusion3,
-        ui_model_type=ModelType.VAE,
+        description=FieldDescriptions.vae_model, ui_type=UIType.VAEModel, title="VAE", default=None
     )
 
     def invoke(self, context: InvocationContext) -> Sd3ModelLoaderOutput:

invokeai/app/invocations/sdxl.py

@@ -1,8 +1,8 @@
 from invokeai.app.invocations.baseinvocation import BaseInvocation, BaseInvocationOutput, invocation, invocation_output
-from invokeai.app.invocations.fields import FieldDescriptions, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, InputField, OutputField, UIType
 from invokeai.app.invocations.model import CLIPField, ModelIdentifierField, UNetField, VAEField
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType, SubModelType
+from invokeai.backend.model_manager.taxonomy import SubModelType
 
 
 @invocation_output("sdxl_model_loader_output")
@@ -29,9 +29,7 @@ class SDXLModelLoaderInvocation(BaseInvocation):
     """Loads an sdxl base model, outputting its submodels."""
 
     model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.sdxl_main_model,
-        ui_model_base=BaseModelType.StableDiffusionXL,
-        ui_model_type=ModelType.Main,
+        description=FieldDescriptions.sdxl_main_model, ui_type=UIType.SDXLMainModel
     )
     # TODO: precision?
 
@@ -69,9 +67,7 @@ class SDXLRefinerModelLoaderInvocation(BaseInvocation):
     """Loads an sdxl refiner model, outputting its submodels."""
 
     model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.sdxl_refiner_model,
-        ui_model_base=BaseModelType.StableDiffusionXLRefiner,
-        ui_model_type=ModelType.Main,
+        description=FieldDescriptions.sdxl_refiner_model, ui_type=UIType.SDXLRefinerModel
     )
     # TODO: precision?
 

invokeai/app/invocations/spandrel_image_to_image.py

@@ -11,6 +11,7 @@ from invokeai.app.invocations.fields import (
     FieldDescriptions,
     ImageField,
     InputField,
+    UIType,
     WithBoard,
     WithMetadata,
 )
@@ -18,7 +19,6 @@ from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.primitives import ImageOutput
 from invokeai.app.services.session_processor.session_processor_common import CanceledException
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.taxonomy import ModelType
 from invokeai.backend.spandrel_image_to_image_model import SpandrelImageToImageModel
 from invokeai.backend.tiles.tiles import calc_tiles_min_overlap
 from invokeai.backend.tiles.utils import TBLR, Tile
@@ -33,7 +33,7 @@ class SpandrelImageToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
     image_to_image_model: ModelIdentifierField = InputField(
         title="Image-to-Image Model",
         description=FieldDescriptions.spandrel_image_to_image_model,
-        ui_model_type=ModelType.SpandrelImageToImage,
+        ui_type=UIType.SpandrelImageToImageModel,
     )
     tile_size: int = InputField(
         default=512, description="The tile size for tiled image-to-image. Set to 0 to disable tiling."

invokeai/app/invocations/t2i_adapter.py

@@ -8,12 +8,11 @@ from invokeai.app.invocations.baseinvocation import (
     invocation,
     invocation_output,
 )
-from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, OutputField
+from invokeai.app.invocations.fields import FieldDescriptions, ImageField, InputField, OutputField, UIType
 from invokeai.app.invocations.model import ModelIdentifierField
 from invokeai.app.invocations.util import validate_begin_end_step, validate_weights
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.controlnet_utils import CONTROLNET_RESIZE_VALUES
-from invokeai.backend.model_manager.taxonomy import BaseModelType, ModelType
 
 
 class T2IAdapterField(BaseModel):
@@ -61,8 +60,7 @@ class T2IAdapterInvocation(BaseInvocation):
         description="The T2I-Adapter model.",
         title="T2I-Adapter Model",
         ui_order=-1,
-        ui_model_base=[BaseModelType.StableDiffusion1, BaseModelType.StableDiffusionXL],
-        ui_model_type=ModelType.T2IAdapter,
+        ui_type=UIType.T2IAdapterModel,
     )
     weight: Union[float, list[float]] = InputField(
         default=1, ge=0, description="The weight given to the T2I-Adapter", title="Weight"

invokeai/backend/ip_adapter/ip_adapter.py

@@ -207,24 +207,15 @@ class IPAdapterPlusXL(IPAdapterPlus):
 
 
 def load_ip_adapter_tensors(ip_adapter_ckpt_path: pathlib.Path, device: str) -> IPAdapterStateDict:
-    state_dict: IPAdapterStateDict = {
-        "ip_adapter": {},
-        "image_proj": {},
-        "adapter_modules": {},  # added for noobai-mark-ipa
-        "image_proj_model": {},  # added for noobai-mark-ipa
-    }
+    state_dict: IPAdapterStateDict = {"ip_adapter": {}, "image_proj": {}}
 
     if ip_adapter_ckpt_path.suffix == ".safetensors":
         model = safetensors.torch.load_file(ip_adapter_ckpt_path, device=device)
         for key in model.keys():
-            if key.startswith("ip_adapter."):
-                state_dict["ip_adapter"][key.replace("ip_adapter.", "")] = model[key]
-            elif key.startswith("image_proj_model."):
-                state_dict["image_proj_model"][key.replace("image_proj_model.", "")] = model[key]
-            elif key.startswith("image_proj."):
+            if key.startswith("image_proj."):
                 state_dict["image_proj"][key.replace("image_proj.", "")] = model[key]
-            elif key.startswith("adapter_modules."):
-                state_dict["adapter_modules"][key.replace("adapter_modules.", "")] = model[key]
+            elif key.startswith("ip_adapter."):
+                state_dict["ip_adapter"][key.replace("ip_adapter.", "")] = model[key]
             else:
                 raise RuntimeError(f"Encountered unexpected IP Adapter state dict key: '{key}'.")
     else:

invokeai/backend/model_manager/config.py

@@ -651,8 +651,6 @@ class LlavaOnevisionConfig(DiffusersConfigBase, ModelConfigBase):
         }
 
 
-
-
 class ApiModelConfig(MainConfigBase, ModelConfigBase):
     """Model config for API-based models."""
 

invokeai/backend/model_manager/load/model_loaders/qwen_image.py

@@ -1,108 +0,0 @@
-# Copyright (c) 2024, Brandon W. Rising and the InvokeAI Development Team
-"""Class for Qwen-Image model loading in InvokeAI."""
-
-from pathlib import Path
-from typing import Optional
-
-from diffusers import DiffusionPipeline
-
-from invokeai.backend.model_manager.config import AnyModelConfig, MainDiffusersConfig
-from invokeai.backend.model_manager.load.load_default import ModelLoader
-from invokeai.backend.model_manager.load.model_loader_registry import ModelLoaderRegistry
-from invokeai.backend.model_manager.load.model_util import calc_model_size_by_fs
-from invokeai.backend.model_manager.taxonomy import (
-    AnyModel,
-    BaseModelType,
-    ModelFormat,
-    ModelType,
-    SubModelType,
-)
-
-
-@ModelLoaderRegistry.register(base=BaseModelType.QwenImage, type=ModelType.Main, format=ModelFormat.Diffusers)
-class QwenImageLoader(ModelLoader):
-    """Class to load Qwen-Image models."""
-
-    def get_size_fs(
-        self, config: AnyModelConfig, model_path: Path, submodel_type: Optional[SubModelType] = None
-    ) -> int:
-        """Calculate the size of the Qwen-Image model on disk."""
-        if not isinstance(config, MainDiffusersConfig):
-            raise ValueError("Only MainDiffusersConfig models are currently supported here.")
-
-        # For Qwen-Image, we need to calculate the size of the entire model or specific submodels
-        return calc_model_size_by_fs(
-            model_path=model_path,
-            subfolder=submodel_type.value if submodel_type else None,
-            variant=config.repo_variant.value if config.repo_variant else None,
-        )
-
-    def _load_model(
-        self,
-        config: AnyModelConfig,
-        submodel_type: Optional[SubModelType] = None,
-    ) -> AnyModel:
-        if not isinstance(config, MainDiffusersConfig):
-            raise ValueError("Only MainDiffusersConfig models are currently supported here.")
-
-        if config.base != BaseModelType.QwenImage:
-            raise ValueError("This loader only supports Qwen-Image models.")
-
-        model_path = Path(config.path)
-
-        if submodel_type is not None:
-            # Load individual submodel components with memory optimizations
-            import torch
-            from diffusers import QwenImageTransformer2DModel
-            from diffusers.models import AutoencoderKLQwenImage
-
-            # Force bfloat16 for memory efficiency if not already set
-            torch_dtype = self._torch_dtype if self._torch_dtype is not None else torch.bfloat16
-
-            # Load only the specific submodel, not the entire pipeline
-            if submodel_type == SubModelType.VAE:
-                # Load VAE directly from subfolder
-                vae_path = model_path / "vae"
-                if vae_path.exists():
-                    return AutoencoderKLQwenImage.from_pretrained(
-                        vae_path,
-                        torch_dtype=torch_dtype,
-                        low_cpu_mem_usage=True,
-                    )
-            elif submodel_type == SubModelType.Transformer:
-                # Load transformer directly from subfolder
-                transformer_path = model_path / "transformer"
-                if transformer_path.exists():
-                    return QwenImageTransformer2DModel.from_pretrained(
-                        transformer_path,
-                        torch_dtype=torch_dtype,
-                        low_cpu_mem_usage=True,
-                    )
-
-            # Fallback to loading full pipeline if direct loading fails
-            pipeline = DiffusionPipeline.from_pretrained(
-                model_path,
-                torch_dtype=torch_dtype,
-                variant=config.repo_variant.value if config.repo_variant else None,
-                low_cpu_mem_usage=True,
-            )
-
-            # Return the specific submodel
-            if hasattr(pipeline, submodel_type.value):
-                return getattr(pipeline, submodel_type.value)
-            else:
-                raise ValueError(f"Submodel {submodel_type} not found in Qwen-Image pipeline.")
-        else:
-            # Load the full pipeline with memory optimizations
-            import torch
-
-            # Force bfloat16 for memory efficiency if not already set
-            torch_dtype = self._torch_dtype if self._torch_dtype is not None else torch.bfloat16
-
-            pipeline = DiffusionPipeline.from_pretrained(
-                model_path,
-                torch_dtype=torch_dtype,
-                variant=config.repo_variant.value if config.repo_variant else None,
-                low_cpu_mem_usage=True,  # Important for reducing memory during loading
-            )
-            return pipeline

invokeai/backend/model_manager/taxonomy.py

@@ -33,7 +33,6 @@ class BaseModelType(str, Enum):
     FluxKontext = "flux-kontext"
     Veo3 = "veo3"
     Runway = "runway"
-    QwenImage = "qwen-image"
 
 
 class ModelType(str, Enum):

invokeai/backend/util/hotfixes.py

@@ -5,7 +5,11 @@ import torch
 from diffusers.configuration_utils import ConfigMixin, register_to_config
 from diffusers.loaders.single_file_model import FromOriginalModelMixin
 from diffusers.models.attention_processor import AttentionProcessor, AttnProcessor
-from diffusers.models.controlnet import ControlNetConditioningEmbedding, ControlNetOutput, zero_module
+from diffusers.models.controlnets.controlnet import (
+    ControlNetConditioningEmbedding,
+    ControlNetOutput,
+    zero_module,
+)
 from diffusers.models.embeddings import (
     TextImageProjection,
     TextImageTimeEmbedding,
@@ -775,7 +779,15 @@ class ControlNetModel(ModelMixin, ConfigMixin, FromOriginalModelMixin):
 
 
 diffusers.ControlNetModel = ControlNetModel
-diffusers.models.controlnet.ControlNetModel = ControlNetModel
+# Patch both the new and legacy module paths for compatibility
+try:
+    diffusers.models.controlnets.controlnet.ControlNetModel = ControlNetModel
+except Exception:
+    # Fallback for environments still exposing the legacy path
+    try:
+        diffusers.models.controlnet.ControlNetModel = ControlNetModel
+    except Exception:
+        pass
 
 
 # patch LoRACompatibleConv to use original Conv2D forward function

invokeai/frontend/web/public/locales/it.json

@@ -131,8 +131,7 @@
         "notInstalled": "Non $t(common.installed)",
         "prevPage": "Pagina precedente",
         "nextPage": "Pagina successiva",
-        "resetToDefaults": "Ripristina impostazioni predefinite",
-        "crop": "Ritaglia"
+        "resetToDefaults": "Ripristina impostazioni predefinite"
     },
     "gallery": {
         "galleryImageSize": "Dimensione dell'immagine",
@@ -279,14 +278,6 @@
             "selectVideoTab": {
                 "title": "Seleziona la scheda Video",
                 "desc": "Seleziona la scheda Video."
-            },
-            "promptHistoryPrev": {
-                "title": "Prompt precedente nella cronologia",
-                "desc": "Quando il prompt è attivo, passa al prompt precedente (più vecchio) nella cronologia."
-            },
-            "promptHistoryNext": {
-                "title": "Prossimo prompt nella cronologia",
-                "desc": "Quando il prompt è attivo, passa al prompt successivo (più recente) nella cronologia."
             }
         },
         "hotkeys": "Tasti di scelta rapida",
@@ -884,8 +875,7 @@
         "video": "Video",
         "resolution": "Risoluzione",
         "downloadImage": "Scarica l'immagine",
-        "showOptionsPanel": "Mostra pannello laterale (O o T)",
-        "startingFrameImageAspectRatioWarning": "Le proporzioni dell'immagine non corrispondono alle proporzioni del video ({{videoAspectRatio}}). Ciò potrebbe causare ritagli imprevisti durante la generazione del video."
+        "showOptionsPanel": "Mostra pannello laterale (O o T)"
     },
     "settings": {
         "models": "Modelli",
@@ -2105,10 +2095,7 @@
         "generateFromImage": "Genera prompt dall'immagine",
         "resultTitle": "Espansione del prompt completata",
         "resultSubtitle": "Scegli come gestire il prompt espanso:",
-        "insert": "Inserisci",
-        "noPromptHistory": "Nessuna cronologia di prompt registrata.",
-        "noMatchingPrompts": "Nessun prompt corrispondente nella cronologia.",
-        "toSwitchBetweenPrompts": "per passare da un prompt all'altro."
+        "insert": "Inserisci"
     },
     "controlLayers": {
         "addLayer": "Aggiungi Livello",
@@ -2804,8 +2791,7 @@
         "watchRecentReleaseVideos": "Guarda i video su questa versione",
         "items": [
             "Seleziona oggetto v2: selezione degli oggetti migliorata con input di punti e riquadri o prompt di testo.",
-            "Regolazioni del livello raster: regola facilmente la luminosità, il contrasto, la saturazione, le curve e altro ancora del livello.",
-            "Cronologia prompt: rivedi e richiama rapidamente i tuoi ultimi 100 prompt."
+            "Regolazioni del livello raster: regola facilmente la luminosità, il contrasto, la saturazione, le curve e altro ancora del livello."
         ],
         "watchUiUpdatesOverview": "Guarda la panoramica degli aggiornamenti dell'interfaccia utente"
     },

invokeai/frontend/web/src/features/modelManagerV2/subpanels/ModelManagerPanel/ModelBaseBadge.tsx

@@ -16,7 +16,6 @@ export const BASE_COLOR_MAP: Record<BaseModelType, string> = {
   'sdxl-refiner': 'invokeBlue',
   flux: 'gold',
   cogview4: 'red',
-  'qwen-image': 'cyan',
   imagen3: 'pink',
   imagen4: 'pink',
   'chatgpt-4o': 'pink',

invokeai/frontend/web/src/features/modelManagerV2/subpanels/ModelPane.tsx

@@ -18,7 +18,6 @@ const modelPaneSx: SystemStyleObject = {
   },
   h: 'full',
   minWidth: '300px',
-  overflow: 'auto',
 };
 
 export const ModelPane = memo(() => {

invokeai/frontend/web/src/features/modelManagerV2/subpanels/ModelPanel/ModelView.tsx

@@ -40,7 +40,7 @@ export const ModelView = memo(({ modelConfig }: Props) => {
   }, [modelConfig.base, modelConfig.type]);
 
   return (
-    <Flex flexDir="column" gap={4} h="full">
+    <Flex flexDir="column" gap={4}>
       <ModelHeader modelConfig={modelConfig}>
         {modelConfig.format === 'checkpoint' && modelConfig.type === 'main' && (
           <ModelConvertButton modelConfig={modelConfig} />
@@ -48,7 +48,7 @@ export const ModelView = memo(({ modelConfig }: Props) => {
         <ModelEditButton />
       </ModelHeader>
       <Divider />
-      <Flex flexDir="column" gap={4}>
+      <Flex flexDir="column" h="full" gap={4}>
         <Box>
           <SimpleGrid columns={2} gap={4}>
             <ModelAttrView label={t('modelManager.baseModel')} value={modelConfig.base} />

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/InputFieldRenderer.tsx

@@ -1,10 +1,14 @@
 import { FloatFieldInput } from 'features/nodes/components/flow/nodes/Invocation/fields/FloatField/FloatFieldInput';
 import { FloatFieldInputAndSlider } from 'features/nodes/components/flow/nodes/Invocation/fields/FloatField/FloatFieldInputAndSlider';
 import { FloatFieldSlider } from 'features/nodes/components/flow/nodes/Invocation/fields/FloatField/FloatFieldSlider';
+import ChatGPT4oModelFieldInputComponent from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ChatGPT4oModelFieldInputComponent';
 import { FloatFieldCollectionInputComponent } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/FloatFieldCollectionInputComponent';
 import { FloatGeneratorFieldInputComponent } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/FloatGeneratorFieldComponent';
+import FluxKontextModelFieldInputComponent from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/FluxKontextModelFieldInputComponent';
 import { ImageFieldCollectionInputComponent } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ImageFieldCollectionInputComponent';
 import { ImageGeneratorFieldInputComponent } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ImageGeneratorFieldComponent';
+import Imagen3ModelFieldInputComponent from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/Imagen3ModelFieldInputComponent';
+import Imagen4ModelFieldInputComponent from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/Imagen4ModelFieldInputComponent';
 import { IntegerFieldCollectionInputComponent } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/IntegerFieldCollectionInputComponent';
 import { IntegerGeneratorFieldInputComponent } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/IntegerGeneratorFieldComponent';
 import ModelIdentifierFieldInputComponent from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelIdentifierFieldInputComponent';
@@ -23,8 +27,22 @@ import {
   isBoardFieldInputTemplate,
   isBooleanFieldInputInstance,
   isBooleanFieldInputTemplate,
+  isChatGPT4oModelFieldInputInstance,
+  isChatGPT4oModelFieldInputTemplate,
+  isCLIPEmbedModelFieldInputInstance,
+  isCLIPEmbedModelFieldInputTemplate,
+  isCLIPGEmbedModelFieldInputInstance,
+  isCLIPGEmbedModelFieldInputTemplate,
+  isCLIPLEmbedModelFieldInputInstance,
+  isCLIPLEmbedModelFieldInputTemplate,
+  isCogView4MainModelFieldInputInstance,
+  isCogView4MainModelFieldInputTemplate,
   isColorFieldInputInstance,
   isColorFieldInputTemplate,
+  isControlLoRAModelFieldInputInstance,
+  isControlLoRAModelFieldInputTemplate,
+  isControlNetModelFieldInputInstance,
+  isControlNetModelFieldInputTemplate,
   isEnumFieldInputInstance,
   isEnumFieldInputTemplate,
   isFloatFieldCollectionInputInstance,
@@ -33,28 +51,68 @@ import {
   isFloatFieldInputTemplate,
   isFloatGeneratorFieldInputInstance,
   isFloatGeneratorFieldInputTemplate,
+  isFluxKontextModelFieldInputInstance,
+  isFluxKontextModelFieldInputTemplate,
+  isFluxMainModelFieldInputInstance,
+  isFluxMainModelFieldInputTemplate,
+  isFluxReduxModelFieldInputInstance,
+  isFluxReduxModelFieldInputTemplate,
+  isFluxVAEModelFieldInputInstance,
+  isFluxVAEModelFieldInputTemplate,
   isImageFieldCollectionInputInstance,
   isImageFieldCollectionInputTemplate,
   isImageFieldInputInstance,
   isImageFieldInputTemplate,
   isImageGeneratorFieldInputInstance,
   isImageGeneratorFieldInputTemplate,
+  isImagen3ModelFieldInputInstance,
+  isImagen3ModelFieldInputTemplate,
+  isImagen4ModelFieldInputInstance,
+  isImagen4ModelFieldInputTemplate,
   isIntegerFieldCollectionInputInstance,
   isIntegerFieldCollectionInputTemplate,
   isIntegerFieldInputInstance,
   isIntegerFieldInputTemplate,
   isIntegerGeneratorFieldInputInstance,
   isIntegerGeneratorFieldInputTemplate,
+  isIPAdapterModelFieldInputInstance,
+  isIPAdapterModelFieldInputTemplate,
+  isLLaVAModelFieldInputInstance,
+  isLLaVAModelFieldInputTemplate,
+  isLoRAModelFieldInputInstance,
+  isLoRAModelFieldInputTemplate,
+  isMainModelFieldInputInstance,
+  isMainModelFieldInputTemplate,
   isModelIdentifierFieldInputInstance,
   isModelIdentifierFieldInputTemplate,
+  isRunwayModelFieldInputInstance,
+  isRunwayModelFieldInputTemplate,
   isSchedulerFieldInputInstance,
   isSchedulerFieldInputTemplate,
+  isSD3MainModelFieldInputInstance,
+  isSD3MainModelFieldInputTemplate,
+  isSDXLMainModelFieldInputInstance,
+  isSDXLMainModelFieldInputTemplate,
+  isSDXLRefinerModelFieldInputInstance,
+  isSDXLRefinerModelFieldInputTemplate,
+  isSigLipModelFieldInputInstance,
+  isSigLipModelFieldInputTemplate,
+  isSpandrelImageToImageModelFieldInputInstance,
+  isSpandrelImageToImageModelFieldInputTemplate,
   isStringFieldCollectionInputInstance,
   isStringFieldCollectionInputTemplate,
   isStringFieldInputInstance,
   isStringFieldInputTemplate,
   isStringGeneratorFieldInputInstance,
   isStringGeneratorFieldInputTemplate,
+  isT2IAdapterModelFieldInputInstance,
+  isT2IAdapterModelFieldInputTemplate,
+  isT5EncoderModelFieldInputInstance,
+  isT5EncoderModelFieldInputTemplate,
+  isVAEModelFieldInputInstance,
+  isVAEModelFieldInputTemplate,
+  isVeo3ModelFieldInputInstance,
+  isVeo3ModelFieldInputTemplate,
 } from 'features/nodes/types/field';
 import type { NodeFieldElement } from 'features/nodes/types/workflow';
 import { memo } from 'react';
@@ -63,10 +121,33 @@ import { assert } from 'tsafe';
 
 import BoardFieldInputComponent from './inputs/BoardFieldInputComponent';
 import BooleanFieldInputComponent from './inputs/BooleanFieldInputComponent';
+import CLIPEmbedModelFieldInputComponent from './inputs/CLIPEmbedModelFieldInputComponent';
+import CLIPGEmbedModelFieldInputComponent from './inputs/CLIPGEmbedModelFieldInputComponent';
+import CLIPLEmbedModelFieldInputComponent from './inputs/CLIPLEmbedModelFieldInputComponent';
+import CogView4MainModelFieldInputComponent from './inputs/CogView4MainModelFieldInputComponent';
 import ColorFieldInputComponent from './inputs/ColorFieldInputComponent';
+import ControlLoRAModelFieldInputComponent from './inputs/ControlLoraModelFieldInputComponent';
+import ControlNetModelFieldInputComponent from './inputs/ControlNetModelFieldInputComponent';
 import EnumFieldInputComponent from './inputs/EnumFieldInputComponent';
+import FluxMainModelFieldInputComponent from './inputs/FluxMainModelFieldInputComponent';
+import FluxReduxModelFieldInputComponent from './inputs/FluxReduxModelFieldInputComponent';
+import FluxVAEModelFieldInputComponent from './inputs/FluxVAEModelFieldInputComponent';
 import ImageFieldInputComponent from './inputs/ImageFieldInputComponent';
+import IPAdapterModelFieldInputComponent from './inputs/IPAdapterModelFieldInputComponent';
+import LLaVAModelFieldInputComponent from './inputs/LLaVAModelFieldInputComponent';
+import LoRAModelFieldInputComponent from './inputs/LoRAModelFieldInputComponent';
+import MainModelFieldInputComponent from './inputs/MainModelFieldInputComponent';
+import RefinerModelFieldInputComponent from './inputs/RefinerModelFieldInputComponent';
+import RunwayModelFieldInputComponent from './inputs/RunwayModelFieldInputComponent';
 import SchedulerFieldInputComponent from './inputs/SchedulerFieldInputComponent';
+import SD3MainModelFieldInputComponent from './inputs/SD3MainModelFieldInputComponent';
+import SDXLMainModelFieldInputComponent from './inputs/SDXLMainModelFieldInputComponent';
+import SigLipModelFieldInputComponent from './inputs/SigLipModelFieldInputComponent';
+import SpandrelImageToImageModelFieldInputComponent from './inputs/SpandrelImageToImageModelFieldInputComponent';
+import T2IAdapterModelFieldInputComponent from './inputs/T2IAdapterModelFieldInputComponent';
+import T5EncoderModelFieldInputComponent from './inputs/T5EncoderModelFieldInputComponent';
+import VAEModelFieldInputComponent from './inputs/VAEModelFieldInputComponent';
+import Veo3ModelFieldInputComponent from './inputs/Veo3ModelFieldInputComponent';
 
 type Props = {
   nodeId: string;
@@ -206,6 +287,13 @@ export const InputFieldRenderer = memo(({ nodeId, fieldName, settings }: Props)
     return <BoardFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
   }
 
+  if (isMainModelFieldInputTemplate(template)) {
+    if (!isMainModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <MainModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
   if (isModelIdentifierFieldInputTemplate(template)) {
     if (!isModelIdentifierFieldInputInstance(field)) {
       return null;
@@ -213,6 +301,159 @@ export const InputFieldRenderer = memo(({ nodeId, fieldName, settings }: Props)
     return <ModelIdentifierFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
   }
 
+  if (isSDXLRefinerModelFieldInputTemplate(template)) {
+    if (!isSDXLRefinerModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <RefinerModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isVAEModelFieldInputTemplate(template)) {
+    if (!isVAEModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <VAEModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isT5EncoderModelFieldInputTemplate(template)) {
+    if (!isT5EncoderModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <T5EncoderModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+  if (isCLIPEmbedModelFieldInputTemplate(template)) {
+    if (!isCLIPEmbedModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <CLIPEmbedModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isCLIPLEmbedModelFieldInputTemplate(template)) {
+    if (!isCLIPLEmbedModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <CLIPLEmbedModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isCLIPGEmbedModelFieldInputTemplate(template)) {
+    if (!isCLIPGEmbedModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <CLIPGEmbedModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isControlLoRAModelFieldInputTemplate(template)) {
+    if (!isControlLoRAModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <ControlLoRAModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isLLaVAModelFieldInputTemplate(template)) {
+    if (!isLLaVAModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <LLaVAModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isFluxVAEModelFieldInputTemplate(template)) {
+    if (!isFluxVAEModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <FluxVAEModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isLoRAModelFieldInputTemplate(template)) {
+    if (!isLoRAModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <LoRAModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isControlNetModelFieldInputTemplate(template)) {
+    if (!isControlNetModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <ControlNetModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isIPAdapterModelFieldInputTemplate(template)) {
+    if (!isIPAdapterModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <IPAdapterModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isT2IAdapterModelFieldInputTemplate(template)) {
+    if (!isT2IAdapterModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <T2IAdapterModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isSpandrelImageToImageModelFieldInputTemplate(template)) {
+    if (!isSpandrelImageToImageModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <SpandrelImageToImageModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isSigLipModelFieldInputTemplate(template)) {
+    if (!isSigLipModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <SigLipModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isFluxReduxModelFieldInputTemplate(template)) {
+    if (!isFluxReduxModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <FluxReduxModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isImagen3ModelFieldInputTemplate(template)) {
+    if (!isImagen3ModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <Imagen3ModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isImagen4ModelFieldInputTemplate(template)) {
+    if (!isImagen4ModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <Imagen4ModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isFluxKontextModelFieldInputTemplate(template)) {
+    if (!isFluxKontextModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <FluxKontextModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isChatGPT4oModelFieldInputTemplate(template)) {
+    if (!isChatGPT4oModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <ChatGPT4oModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isVeo3ModelFieldInputTemplate(template)) {
+    if (!isVeo3ModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <Veo3ModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isRunwayModelFieldInputTemplate(template)) {
+    if (!isRunwayModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <RunwayModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
   if (isColorFieldInputTemplate(template)) {
     if (!isColorFieldInputInstance(field)) {
       return null;
@@ -220,6 +461,34 @@ export const InputFieldRenderer = memo(({ nodeId, fieldName, settings }: Props)
     return <ColorFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
   }
 
+  if (isFluxMainModelFieldInputTemplate(template)) {
+    if (!isFluxMainModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <FluxMainModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isSD3MainModelFieldInputTemplate(template)) {
+    if (!isSD3MainModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <SD3MainModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isCogView4MainModelFieldInputTemplate(template)) {
+    if (!isCogView4MainModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <CogView4MainModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
+  if (isSDXLMainModelFieldInputTemplate(template)) {
+    if (!isSDXLMainModelFieldInputInstance(field)) {
+      return null;
+    }
+    return <SDXLMainModelFieldInputComponent nodeId={nodeId} field={field} fieldTemplate={template} />;
+  }
+
   if (isSchedulerFieldInputTemplate(template)) {
     if (!isSchedulerFieldInputInstance(field)) {
       return null;

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/CLIPEmbedModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldCLIPEmbedValueChanged } from 'features/nodes/store/nodesSlice';
+import type { CLIPEmbedModelFieldInputInstance, CLIPEmbedModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useCLIPEmbedModels } from 'services/api/hooks/modelsByType';
+import type { CLIPEmbedModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<CLIPEmbedModelFieldInputInstance, CLIPEmbedModelFieldInputTemplate>;
+
+const CLIPEmbedModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useCLIPEmbedModels();
+  const onChange = useCallback(
+    (value: CLIPEmbedModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldCLIPEmbedValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(CLIPEmbedModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/CLIPGEmbedModelFieldInputComponent.tsx

@@ -0,0 +1,45 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldCLIPGEmbedValueChanged } from 'features/nodes/store/nodesSlice';
+import type { CLIPGEmbedModelFieldInputInstance, CLIPGEmbedModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useCLIPEmbedModels } from 'services/api/hooks/modelsByType';
+import { type CLIPGEmbedModelConfig, isCLIPGEmbedModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<CLIPGEmbedModelFieldInputInstance, CLIPGEmbedModelFieldInputTemplate>;
+
+const CLIPGEmbedModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useCLIPEmbedModels();
+
+  const onChange = useCallback(
+    (value: CLIPGEmbedModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldCLIPGEmbedValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs.filter((config) => isCLIPGEmbedModelConfig(config))}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(CLIPGEmbedModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/CLIPLEmbedModelFieldInputComponent.tsx

@@ -0,0 +1,45 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldCLIPLEmbedValueChanged } from 'features/nodes/store/nodesSlice';
+import type { CLIPLEmbedModelFieldInputInstance, CLIPLEmbedModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useCLIPEmbedModels } from 'services/api/hooks/modelsByType';
+import { type CLIPLEmbedModelConfig, isCLIPLEmbedModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<CLIPLEmbedModelFieldInputInstance, CLIPLEmbedModelFieldInputTemplate>;
+
+const CLIPLEmbedModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useCLIPEmbedModels();
+
+  const onChange = useCallback(
+    (value: CLIPLEmbedModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldCLIPLEmbedValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs.filter((config) => isCLIPLEmbedModelConfig(config))}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(CLIPLEmbedModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/ChatGPT4oModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldChatGPT4oModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { ChatGPT4oModelFieldInputInstance, ChatGPT4oModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useChatGPT4oModels } from 'services/api/hooks/modelsByType';
+import type { ApiModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const ChatGPT4oModelFieldInputComponent = (
+  props: FieldComponentProps<ChatGPT4oModelFieldInputInstance, ChatGPT4oModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useChatGPT4oModels();
+
+  const onChange = useCallback(
+    (value: ApiModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldChatGPT4oModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(ChatGPT4oModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/CogView4MainModelFieldInputComponent.tsx

@@ -0,0 +1,63 @@
+import { Combobox, Flex, FormControl } from '@invoke-ai/ui-library';
+import { useAppDispatch } from 'app/store/storeHooks';
+import { useGroupedModelCombobox } from 'common/hooks/useGroupedModelCombobox';
+import { fieldMainModelValueChanged } from 'features/nodes/store/nodesSlice';
+import { NO_DRAG_CLASS, NO_WHEEL_CLASS } from 'features/nodes/types/constants';
+import type {
+  CogView4MainModelFieldInputInstance,
+  CogView4MainModelFieldInputTemplate,
+} from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useCogView4Models } from 'services/api/hooks/modelsByType';
+import type { MainModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<CogView4MainModelFieldInputInstance, CogView4MainModelFieldInputTemplate>;
+
+const CogView4MainModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useCogView4Models();
+  const _onChange = useCallback(
+    (value: MainModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldMainModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+  const { options, value, onChange, placeholder, noOptionsMessage } = useGroupedModelCombobox({
+    modelConfigs,
+    onChange: _onChange,
+    isLoading,
+    selectedModel: field.value,
+  });
+
+  return (
+    <Flex w="full" alignItems="center" gap={2}>
+      <FormControl
+        className={`${NO_WHEEL_CLASS} ${NO_DRAG_CLASS}`}
+        isDisabled={!options.length}
+        isInvalid={!value && props.fieldTemplate.required}
+      >
+        <Combobox
+          value={value}
+          placeholder={placeholder}
+          options={options}
+          onChange={onChange}
+          noOptionsMessage={noOptionsMessage}
+        />
+      </FormControl>
+    </Flex>
+  );
+};
+
+export default memo(CogView4MainModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/ControlLoraModelFieldInputComponent.tsx

@@ -0,0 +1,48 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldControlLoRAModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type {
+  ControlLoRAModelFieldInputInstance,
+  ControlLoRAModelFieldInputTemplate,
+} from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useControlLoRAModel } from 'services/api/hooks/modelsByType';
+import type { ControlLoRAModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<ControlLoRAModelFieldInputInstance, ControlLoRAModelFieldInputTemplate>;
+
+const ControlLoRAModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useControlLoRAModel();
+
+  const onChange = useCallback(
+    (value: ControlLoRAModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldControlLoRAModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(ControlLoRAModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/ControlNetModelFieldInputComponent.tsx

@@ -0,0 +1,45 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldControlNetModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { ControlNetModelFieldInputInstance, ControlNetModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useControlNetModels } from 'services/api/hooks/modelsByType';
+import type { ControlNetModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<ControlNetModelFieldInputInstance, ControlNetModelFieldInputTemplate>;
+
+const ControlNetModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useControlNetModels();
+
+  const onChange = useCallback(
+    (value: ControlNetModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldControlNetModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(ControlNetModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/FluxKontextModelFieldInputComponent.tsx

@@ -0,0 +1,49 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldFluxKontextModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type {
+  FluxKontextModelFieldInputInstance,
+  FluxKontextModelFieldInputTemplate,
+} from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useFluxKontextModels } from 'services/api/hooks/modelsByType';
+import type { ApiModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const FluxKontextModelFieldInputComponent = (
+  props: FieldComponentProps<FluxKontextModelFieldInputInstance, FluxKontextModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useFluxKontextModels();
+
+  const onChange = useCallback(
+    (value: ApiModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldFluxKontextModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(FluxKontextModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/FluxMainModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldMainModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { FluxMainModelFieldInputInstance, FluxMainModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useFluxModels } from 'services/api/hooks/modelsByType';
+import type { MainModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<FluxMainModelFieldInputInstance, FluxMainModelFieldInputTemplate>;
+
+const FluxMainModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useFluxModels();
+  const onChange = useCallback(
+    (value: MainModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldMainModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(FluxMainModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/FluxReduxModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldFluxReduxModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { FluxReduxModelFieldInputInstance, FluxReduxModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useFluxReduxModels } from 'services/api/hooks/modelsByType';
+import type { FLUXReduxModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const FluxReduxModelFieldInputComponent = (
+  props: FieldComponentProps<FluxReduxModelFieldInputInstance, FluxReduxModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useFluxReduxModels();
+
+  const onChange = useCallback(
+    (value: FLUXReduxModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldFluxReduxModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(FluxReduxModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/FluxVAEModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldFluxVAEModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { FluxVAEModelFieldInputInstance, FluxVAEModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useFluxVAEModels } from 'services/api/hooks/modelsByType';
+import type { VAEModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<FluxVAEModelFieldInputInstance, FluxVAEModelFieldInputTemplate>;
+
+const FluxVAEModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useFluxVAEModels();
+  const onChange = useCallback(
+    (value: VAEModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldFluxVAEModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(FluxVAEModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/IPAdapterModelFieldInputComponent.tsx

@@ -0,0 +1,45 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldIPAdapterModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { IPAdapterModelFieldInputInstance, IPAdapterModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useIPAdapterModels } from 'services/api/hooks/modelsByType';
+import type { IPAdapterModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const IPAdapterModelFieldInputComponent = (
+  props: FieldComponentProps<IPAdapterModelFieldInputInstance, IPAdapterModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useIPAdapterModels();
+
+  const onChange = useCallback(
+    (value: IPAdapterModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldIPAdapterModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(IPAdapterModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/Imagen3ModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldImagen3ModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { Imagen3ModelFieldInputInstance, Imagen3ModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useImagen3Models } from 'services/api/hooks/modelsByType';
+import type { ApiModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const Imagen3ModelFieldInputComponent = (
+  props: FieldComponentProps<Imagen3ModelFieldInputInstance, Imagen3ModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useImagen3Models();
+
+  const onChange = useCallback(
+    (value: ApiModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldImagen3ModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(Imagen3ModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/Imagen4ModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldImagen4ModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { Imagen4ModelFieldInputInstance, Imagen4ModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useImagen4Models } from 'services/api/hooks/modelsByType';
+import type { ApiModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const Imagen4ModelFieldInputComponent = (
+  props: FieldComponentProps<Imagen4ModelFieldInputInstance, Imagen4ModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useImagen4Models();
+
+  const onChange = useCallback(
+    (value: ApiModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldImagen4ModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(Imagen4ModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/LLaVAModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldLLaVAModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { LLaVAModelFieldInputInstance, LLaVAModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useLLaVAModels } from 'services/api/hooks/modelsByType';
+import type { LlavaOnevisionConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<LLaVAModelFieldInputInstance, LLaVAModelFieldInputTemplate>;
+
+const LLaVAModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useLLaVAModels();
+  const onChange = useCallback(
+    (value: LlavaOnevisionConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldLLaVAModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(LLaVAModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/LoRAModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldLoRAModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { LoRAModelFieldInputInstance, LoRAModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useLoRAModels } from 'services/api/hooks/modelsByType';
+import type { LoRAModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<LoRAModelFieldInputInstance, LoRAModelFieldInputTemplate>;
+
+const LoRAModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useLoRAModels();
+  const onChange = useCallback(
+    (value: LoRAModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldLoRAModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(LoRAModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/MainModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldMainModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { MainModelFieldInputInstance, MainModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useNonSDXLMainModels } from 'services/api/hooks/modelsByType';
+import type { MainModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<MainModelFieldInputInstance, MainModelFieldInputTemplate>;
+
+const MainModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useNonSDXLMainModels();
+  const onChange = useCallback(
+    (value: MainModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldMainModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(MainModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelIdentifierFieldInputComponent.tsx

@@ -12,7 +12,7 @@ import type { FieldComponentProps } from './types';
 type Props = FieldComponentProps<ModelIdentifierFieldInputInstance, ModelIdentifierFieldInputTemplate>;
 
 const ModelIdentifierFieldInputComponent = (props: Props) => {
-  const { nodeId, field, fieldTemplate } = props;
+  const { nodeId, field } = props;
   const dispatch = useAppDispatch();
   const { data, isLoading } = useGetModelConfigsQuery();
   const onChange = useCallback(
@@ -36,31 +36,8 @@ const ModelIdentifierFieldInputComponent = (props: Props) => {
       return EMPTY_ARRAY;
     }
 
-    if (!fieldTemplate.ui_model_base && !fieldTemplate.ui_model_type) {
-      return modelConfigsAdapterSelectors.selectAll(data);
-    }
-
-    return modelConfigsAdapterSelectors.selectAll(data).filter((config) => {
-      if (fieldTemplate.ui_model_base && !fieldTemplate.ui_model_base.includes(config.base)) {
-        return false;
-      }
-      if (fieldTemplate.ui_model_type && !fieldTemplate.ui_model_type.includes(config.type)) {
-        return false;
-      }
-      if (
-        fieldTemplate.ui_model_variant &&
-        'variant' in config &&
-        config.variant &&
-        !fieldTemplate.ui_model_variant.includes(config.variant)
-      ) {
-        return false;
-      }
-      if (fieldTemplate.ui_model_format && !fieldTemplate.ui_model_format.includes(config.format)) {
-        return false;
-      }
-      return true;
-    });
-  }, [data, fieldTemplate]);
+    return modelConfigsAdapterSelectors.selectAll(data);
+  }, [data]);
 
   return (
     <ModelFieldCombobox

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/RefinerModelFieldInputComponent.tsx

@@ -0,0 +1,47 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldRefinerModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type {
+  SDXLRefinerModelFieldInputInstance,
+  SDXLRefinerModelFieldInputTemplate,
+} from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useRefinerModels } from 'services/api/hooks/modelsByType';
+import type { MainModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<SDXLRefinerModelFieldInputInstance, SDXLRefinerModelFieldInputTemplate>;
+
+const RefinerModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useRefinerModels();
+  const onChange = useCallback(
+    (value: MainModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldRefinerModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(RefinerModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/RunwayModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldRunwayModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { RunwayModelFieldInputInstance, RunwayModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useRunwayModels } from 'services/api/hooks/modelsByType';
+import type { VideoApiModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const RunwayModelFieldInputComponent = (
+  props: FieldComponentProps<RunwayModelFieldInputInstance, RunwayModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useRunwayModels();
+
+  const onChange = useCallback(
+    (value: VideoApiModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldRunwayModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(RunwayModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/SD3MainModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldMainModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { SD3MainModelFieldInputInstance, SD3MainModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useSD3Models } from 'services/api/hooks/modelsByType';
+import type { MainModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<SD3MainModelFieldInputInstance, SD3MainModelFieldInputTemplate>;
+
+const SD3MainModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useSD3Models();
+  const onChange = useCallback(
+    (value: MainModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldMainModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(SD3MainModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/SDXLMainModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldMainModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { SDXLMainModelFieldInputInstance, SDXLMainModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useSDXLModels } from 'services/api/hooks/modelsByType';
+import type { MainModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<SDXLMainModelFieldInputInstance, SDXLMainModelFieldInputTemplate>;
+
+const SDXLMainModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useSDXLModels();
+  const onChange = useCallback(
+    (value: MainModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldMainModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(SDXLMainModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/SigLipModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldSigLipModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { SigLipModelFieldInputInstance, SigLipModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useSigLipModels } from 'services/api/hooks/modelsByType';
+import type { SigLipModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const SigLipModelFieldInputComponent = (
+  props: FieldComponentProps<SigLipModelFieldInputInstance, SigLipModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useSigLipModels();
+
+  const onChange = useCallback(
+    (value: SigLipModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldSigLipModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(SigLipModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/SpandrelImageToImageModelFieldInputComponent.tsx

@@ -0,0 +1,49 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldSpandrelImageToImageModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type {
+  SpandrelImageToImageModelFieldInputInstance,
+  SpandrelImageToImageModelFieldInputTemplate,
+} from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useSpandrelImageToImageModels } from 'services/api/hooks/modelsByType';
+import type { SpandrelImageToImageModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const SpandrelImageToImageModelFieldInputComponent = (
+  props: FieldComponentProps<SpandrelImageToImageModelFieldInputInstance, SpandrelImageToImageModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useSpandrelImageToImageModels();
+
+  const onChange = useCallback(
+    (value: SpandrelImageToImageModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldSpandrelImageToImageModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(SpandrelImageToImageModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/T2IAdapterModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldT2IAdapterModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { T2IAdapterModelFieldInputInstance, T2IAdapterModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useT2IAdapterModels } from 'services/api/hooks/modelsByType';
+import type { T2IAdapterModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const T2IAdapterModelFieldInputComponent = (
+  props: FieldComponentProps<T2IAdapterModelFieldInputInstance, T2IAdapterModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useT2IAdapterModels();
+
+  const onChange = useCallback(
+    (value: T2IAdapterModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldT2IAdapterModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(T2IAdapterModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/T5EncoderModelFieldInputComponent.tsx

@@ -0,0 +1,43 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldT5EncoderValueChanged } from 'features/nodes/store/nodesSlice';
+import type { T5EncoderModelFieldInputInstance, T5EncoderModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useT5EncoderModels } from 'services/api/hooks/modelsByType';
+import type { T5EncoderBnbQuantizedLlmInt8bModelConfig, T5EncoderModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<T5EncoderModelFieldInputInstance, T5EncoderModelFieldInputTemplate>;
+
+const T5EncoderModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useT5EncoderModels();
+  const onChange = useCallback(
+    (value: T5EncoderBnbQuantizedLlmInt8bModelConfig | T5EncoderModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldT5EncoderValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(T5EncoderModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/VAEModelFieldInputComponent.tsx

@@ -0,0 +1,44 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldVaeModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { VAEModelFieldInputInstance, VAEModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useVAEModels } from 'services/api/hooks/modelsByType';
+import type { VAEModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+type Props = FieldComponentProps<VAEModelFieldInputInstance, VAEModelFieldInputTemplate>;
+
+const VAEModelFieldInputComponent = (props: Props) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+  const [modelConfigs, { isLoading }] = useVAEModels();
+  const onChange = useCallback(
+    (value: VAEModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldVaeModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(VAEModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/components/flow/nodes/Invocation/fields/inputs/Veo3ModelFieldInputComponent.tsx

@@ -0,0 +1,46 @@
+import { useAppDispatch } from 'app/store/storeHooks';
+import { ModelFieldCombobox } from 'features/nodes/components/flow/nodes/Invocation/fields/inputs/ModelFieldCombobox';
+import { fieldVeo3ModelValueChanged } from 'features/nodes/store/nodesSlice';
+import type { Veo3ModelFieldInputInstance, Veo3ModelFieldInputTemplate } from 'features/nodes/types/field';
+import { memo, useCallback } from 'react';
+import { useVeo3Models } from 'services/api/hooks/modelsByType';
+import type { VideoApiModelConfig } from 'services/api/types';
+
+import type { FieldComponentProps } from './types';
+
+const Veo3ModelFieldInputComponent = (
+  props: FieldComponentProps<Veo3ModelFieldInputInstance, Veo3ModelFieldInputTemplate>
+) => {
+  const { nodeId, field } = props;
+  const dispatch = useAppDispatch();
+
+  const [modelConfigs, { isLoading }] = useVeo3Models();
+
+  const onChange = useCallback(
+    (value: VideoApiModelConfig | null) => {
+      if (!value) {
+        return;
+      }
+      dispatch(
+        fieldVeo3ModelValueChanged({
+          nodeId,
+          fieldName: field.name,
+          value,
+        })
+      );
+    },
+    [dispatch, field.name, nodeId]
+  );
+
+  return (
+    <ModelFieldCombobox
+      value={field.value}
+      modelConfigs={modelConfigs}
+      isLoadingConfigs={isLoading}
+      onChange={onChange}
+      required={props.fieldTemplate.required}
+    />
+  );
+};
+
+export default memo(Veo3ModelFieldInputComponent);

invokeai/frontend/web/src/features/nodes/hooks/useInputFieldNamesByStatus.ts

@@ -2,16 +2,20 @@ import { createSelector } from '@reduxjs/toolkit';
 import { useAppSelector } from 'app/store/storeHooks';
 import { useInvocationNodeContext } from 'features/nodes/components/flow/nodes/Invocation/context';
 import type { FieldInputTemplate } from 'features/nodes/types/field';
-import { isSingleOrCollection, isStatefulFieldType } from 'features/nodes/types/field';
+import { isSingleOrCollection } from 'features/nodes/types/field';
+import { TEMPLATE_BUILDER_MAP } from 'features/nodes/util/schema/buildFieldInputTemplate';
 import { useMemo } from 'react';
 
 const isConnectionInputField = (field: FieldInputTemplate) => {
-  return (field.input === 'connection' && !isSingleOrCollection(field.type)) || !isStatefulFieldType(field.type);
+  return (
+    (field.input === 'connection' && !isSingleOrCollection(field.type)) || !(field.type.name in TEMPLATE_BUILDER_MAP)
+  );
 };
 
 const isAnyOrDirectInputField = (field: FieldInputTemplate) => {
   return (
-    (['any', 'direct'].includes(field.input) || isSingleOrCollection(field.type)) && isStatefulFieldType(field.type)
+    (['any', 'direct'].includes(field.input) || isSingleOrCollection(field.type)) &&
+    field.type.name in TEMPLATE_BUILDER_MAP
   );
 };
 

invokeai/frontend/web/src/features/nodes/store/nodesSlice.ts

@@ -24,44 +24,90 @@ import { SHARED_NODE_PROPERTIES } from 'features/nodes/types/constants';
 import type {
   BoardFieldValue,
   BooleanFieldValue,
+  ChatGPT4oModelFieldValue,
+  CLIPEmbedModelFieldValue,
+  CLIPGEmbedModelFieldValue,
+  CLIPLEmbedModelFieldValue,
   ColorFieldValue,
+  ControlLoRAModelFieldValue,
+  ControlNetModelFieldValue,
   EnumFieldValue,
   FieldValue,
   FloatFieldValue,
   FloatGeneratorFieldValue,
+  FluxKontextModelFieldValue,
+  FluxReduxModelFieldValue,
+  FluxVAEModelFieldValue,
   ImageFieldCollectionValue,
   ImageFieldValue,
   ImageGeneratorFieldValue,
+  Imagen3ModelFieldValue,
+  Imagen4ModelFieldValue,
   IntegerFieldCollectionValue,
   IntegerFieldValue,
   IntegerGeneratorFieldValue,
+  IPAdapterModelFieldValue,
+  LLaVAModelFieldValue,
+  LoRAModelFieldValue,
+  MainModelFieldValue,
   ModelIdentifierFieldValue,
+  RunwayModelFieldValue,
   SchedulerFieldValue,
+  SDXLRefinerModelFieldValue,
+  SigLipModelFieldValue,
+  SpandrelImageToImageModelFieldValue,
   StatefulFieldValue,
   StringFieldCollectionValue,
   StringFieldValue,
   StringGeneratorFieldValue,
+  T2IAdapterModelFieldValue,
+  T5EncoderModelFieldValue,
+  VAEModelFieldValue,
+  Veo3ModelFieldValue,
 } from 'features/nodes/types/field';
 import {
   zBoardFieldValue,
   zBooleanFieldValue,
+  zChatGPT4oModelFieldValue,
+  zCLIPEmbedModelFieldValue,
+  zCLIPGEmbedModelFieldValue,
+  zCLIPLEmbedModelFieldValue,
   zColorFieldValue,
+  zControlLoRAModelFieldValue,
+  zControlNetModelFieldValue,
   zEnumFieldValue,
   zFloatFieldCollectionValue,
   zFloatFieldValue,
   zFloatGeneratorFieldValue,
+  zFluxKontextModelFieldValue,
+  zFluxReduxModelFieldValue,
+  zFluxVAEModelFieldValue,
   zImageFieldCollectionValue,
   zImageFieldValue,
   zImageGeneratorFieldValue,
+  zImagen3ModelFieldValue,
+  zImagen4ModelFieldValue,
   zIntegerFieldCollectionValue,
   zIntegerFieldValue,
   zIntegerGeneratorFieldValue,
+  zIPAdapterModelFieldValue,
+  zLLaVAModelFieldValue,
+  zLoRAModelFieldValue,
+  zMainModelFieldValue,
   zModelIdentifierFieldValue,
+  zRunwayModelFieldValue,
   zSchedulerFieldValue,
+  zSDXLRefinerModelFieldValue,
+  zSigLipModelFieldValue,
+  zSpandrelImageToImageModelFieldValue,
   zStatefulFieldValue,
   zStringFieldCollectionValue,
   zStringFieldValue,
   zStringGeneratorFieldValue,
+  zT2IAdapterModelFieldValue,
+  zT5EncoderModelFieldValue,
+  zVAEModelFieldValue,
+  zVeo3ModelFieldValue,
 } from 'features/nodes/types/field';
 import type { AnyEdge, AnyNode } from 'features/nodes/types/invocation';
 import { isInvocationNode, isNotesNode } from 'features/nodes/types/invocation';
@@ -447,9 +493,81 @@ const slice = createSlice({
     fieldColorValueChanged: (state, action: FieldValueAction<ColorFieldValue>) => {
       fieldValueReducer(state, action, zColorFieldValue);
     },
+    fieldMainModelValueChanged: (state, action: FieldValueAction<MainModelFieldValue>) => {
+      fieldValueReducer(state, action, zMainModelFieldValue);
+    },
     fieldModelIdentifierValueChanged: (state, action: FieldValueAction<ModelIdentifierFieldValue>) => {
       fieldValueReducer(state, action, zModelIdentifierFieldValue);
     },
+    fieldRefinerModelValueChanged: (state, action: FieldValueAction<SDXLRefinerModelFieldValue>) => {
+      fieldValueReducer(state, action, zSDXLRefinerModelFieldValue);
+    },
+    fieldVaeModelValueChanged: (state, action: FieldValueAction<VAEModelFieldValue>) => {
+      fieldValueReducer(state, action, zVAEModelFieldValue);
+    },
+    fieldLoRAModelValueChanged: (state, action: FieldValueAction<LoRAModelFieldValue>) => {
+      fieldValueReducer(state, action, zLoRAModelFieldValue);
+    },
+    fieldLLaVAModelValueChanged: (state, action: FieldValueAction<LLaVAModelFieldValue>) => {
+      fieldValueReducer(state, action, zLLaVAModelFieldValue);
+    },
+    fieldControlNetModelValueChanged: (state, action: FieldValueAction<ControlNetModelFieldValue>) => {
+      fieldValueReducer(state, action, zControlNetModelFieldValue);
+    },
+    fieldIPAdapterModelValueChanged: (state, action: FieldValueAction<IPAdapterModelFieldValue>) => {
+      fieldValueReducer(state, action, zIPAdapterModelFieldValue);
+    },
+    fieldT2IAdapterModelValueChanged: (state, action: FieldValueAction<T2IAdapterModelFieldValue>) => {
+      fieldValueReducer(state, action, zT2IAdapterModelFieldValue);
+    },
+    fieldSpandrelImageToImageModelValueChanged: (
+      state,
+      action: FieldValueAction<SpandrelImageToImageModelFieldValue>
+    ) => {
+      fieldValueReducer(state, action, zSpandrelImageToImageModelFieldValue);
+    },
+    fieldT5EncoderValueChanged: (state, action: FieldValueAction<T5EncoderModelFieldValue>) => {
+      fieldValueReducer(state, action, zT5EncoderModelFieldValue);
+    },
+    fieldCLIPEmbedValueChanged: (state, action: FieldValueAction<CLIPEmbedModelFieldValue>) => {
+      fieldValueReducer(state, action, zCLIPEmbedModelFieldValue);
+    },
+    fieldCLIPLEmbedValueChanged: (state, action: FieldValueAction<CLIPLEmbedModelFieldValue>) => {
+      fieldValueReducer(state, action, zCLIPLEmbedModelFieldValue);
+    },
+    fieldCLIPGEmbedValueChanged: (state, action: FieldValueAction<CLIPGEmbedModelFieldValue>) => {
+      fieldValueReducer(state, action, zCLIPGEmbedModelFieldValue);
+    },
+    fieldControlLoRAModelValueChanged: (state, action: FieldValueAction<ControlLoRAModelFieldValue>) => {
+      fieldValueReducer(state, action, zControlLoRAModelFieldValue);
+    },
+    fieldFluxVAEModelValueChanged: (state, action: FieldValueAction<FluxVAEModelFieldValue>) => {
+      fieldValueReducer(state, action, zFluxVAEModelFieldValue);
+    },
+    fieldSigLipModelValueChanged: (state, action: FieldValueAction<SigLipModelFieldValue>) => {
+      fieldValueReducer(state, action, zSigLipModelFieldValue);
+    },
+    fieldFluxReduxModelValueChanged: (state, action: FieldValueAction<FluxReduxModelFieldValue>) => {
+      fieldValueReducer(state, action, zFluxReduxModelFieldValue);
+    },
+    fieldImagen3ModelValueChanged: (state, action: FieldValueAction<Imagen3ModelFieldValue>) => {
+      fieldValueReducer(state, action, zImagen3ModelFieldValue);
+    },
+    fieldImagen4ModelValueChanged: (state, action: FieldValueAction<Imagen4ModelFieldValue>) => {
+      fieldValueReducer(state, action, zImagen4ModelFieldValue);
+    },
+    fieldChatGPT4oModelValueChanged: (state, action: FieldValueAction<ChatGPT4oModelFieldValue>) => {
+      fieldValueReducer(state, action, zChatGPT4oModelFieldValue);
+    },
+    fieldVeo3ModelValueChanged: (state, action: FieldValueAction<Veo3ModelFieldValue>) => {
+      fieldValueReducer(state, action, zVeo3ModelFieldValue);
+    },
+    fieldRunwayModelValueChanged: (state, action: FieldValueAction<RunwayModelFieldValue>) => {
+      fieldValueReducer(state, action, zRunwayModelFieldValue);
+    },
+    fieldFluxKontextModelValueChanged: (state, action: FieldValueAction<FluxKontextModelFieldValue>) => {
+      fieldValueReducer(state, action, zFluxKontextModelFieldValue);
+    },
     fieldEnumModelValueChanged: (state, action: FieldValueAction<EnumFieldValue>) => {
       fieldValueReducer(state, action, zEnumFieldValue);
     },
@@ -588,22 +706,45 @@ export const {
   fieldBoardValueChanged,
   fieldBooleanValueChanged,
   fieldColorValueChanged,
+  fieldControlNetModelValueChanged,
   fieldEnumModelValueChanged,
   fieldImageValueChanged,
   fieldImageCollectionValueChanged,
+  fieldIPAdapterModelValueChanged,
+  fieldT2IAdapterModelValueChanged,
+  fieldSpandrelImageToImageModelValueChanged,
   fieldLabelChanged,
+  fieldLoRAModelValueChanged,
+  fieldLLaVAModelValueChanged,
   fieldModelIdentifierValueChanged,
+  fieldMainModelValueChanged,
   fieldIntegerValueChanged,
   fieldFloatValueChanged,
   fieldFloatCollectionValueChanged,
   fieldIntegerCollectionValueChanged,
+  fieldRefinerModelValueChanged,
   fieldSchedulerValueChanged,
   fieldStringValueChanged,
   fieldStringCollectionValueChanged,
+  fieldVaeModelValueChanged,
+  fieldT5EncoderValueChanged,
+  fieldCLIPEmbedValueChanged,
+  fieldCLIPLEmbedValueChanged,
+  fieldCLIPGEmbedValueChanged,
+  fieldControlLoRAModelValueChanged,
+  fieldFluxVAEModelValueChanged,
+  fieldSigLipModelValueChanged,
+  fieldFluxReduxModelValueChanged,
+  fieldImagen3ModelValueChanged,
+  fieldImagen4ModelValueChanged,
+  fieldChatGPT4oModelValueChanged,
+  fieldFluxKontextModelValueChanged,
   fieldFloatGeneratorValueChanged,
   fieldIntegerGeneratorValueChanged,
   fieldStringGeneratorValueChanged,
   fieldImageGeneratorValueChanged,
+  fieldVeo3ModelValueChanged,
+  fieldRunwayModelValueChanged,
   fieldDescriptionChanged,
   nodeEditorReset,
   nodeIsIntermediateChanged,

invokeai/frontend/web/src/features/nodes/store/util/areTypesEqual.test.ts

@@ -40,7 +40,7 @@ describe(areTypesEqual.name, () => {
       },
     };
     const targetType: FieldType = {
-      name: 'StringField',
+      name: 'MainModelField',
       cardinality: 'SINGLE',
       batch: false,
       originalType: {
@@ -54,7 +54,7 @@ describe(areTypesEqual.name, () => {
 
   it('should handle equal original source type and target type', () => {
     const sourceType: FieldType = {
-      name: 'FloatField',
+      name: 'MainModelField',
       cardinality: 'SINGLE',
       batch: false,
       originalType: {
@@ -78,7 +78,7 @@ describe(areTypesEqual.name, () => {
 
   it('should handle equal original source type and original target type', () => {
     const sourceType: FieldType = {
-      name: 'IntegerField',
+      name: 'MainModelField',
       cardinality: 'SINGLE',
       batch: false,
       originalType: {
@@ -88,7 +88,7 @@ describe(areTypesEqual.name, () => {
       },
     };
     const targetType: FieldType = {
-      name: 'StringField',
+      name: 'LoRAModelField',
       cardinality: 'SINGLE',
       batch: false,
       originalType: {

invokeai/frontend/web/src/features/nodes/store/util/testUtils.ts

@@ -247,12 +247,17 @@ export const main_model_loader: InvocationTemplate = {
       fieldKind: 'input',
       input: 'direct',
       ui_hidden: false,
-      ui_model_base: ['sd-1'],
-      ui_model_type: ['main'],
+      ui_type: 'MainModelField',
       type: {
-        name: 'ModelIdentifierField',
+        name: 'MainModelField',
         cardinality: 'SINGLE',
         batch: false,
+
+        originalType: {
+          name: 'ModelIdentifierField',
+          cardinality: 'SINGLE',
+          batch: false,
+        },
       },
     },
   },
@@ -791,8 +796,7 @@ export const schema = {
             input: 'direct',
             orig_required: true,
             ui_hidden: false,
-            ui_model_base: ['sd-1'],
-            ui_model_type: ['main'],
+            ui_type: 'MainModelField',
           },
           type: {
             type: 'string',

invokeai/frontend/web/src/features/nodes/types/common.test-d.ts

@@ -12,15 +12,11 @@ import type {
   SchedulerField,
   SubModelType,
   T2IAdapterField,
-  zClipVariantType,
-  zModelFormat,
-  zModelVariantType,
 } from 'features/nodes/types/common';
-import type { Invocation, S } from 'services/api/types';
+import type { Invocation, ModelType, S } from 'services/api/types';
 import type { Equals, Extends } from 'tsafe';
 import { assert } from 'tsafe';
 import { describe, test } from 'vitest';
-import type z from 'zod';
 
 /**
  * These types originate from the server and are recreated as zod schemas manually, for use at runtime.
@@ -42,9 +38,7 @@ describe('Common types', () => {
   test('ModelIdentifier', () => assert<Equals<ModelIdentifierField, S['ModelIdentifierField']>>());
   test('ModelIdentifier', () => assert<Equals<BaseModelType, S['BaseModelType']>>());
   test('ModelIdentifier', () => assert<Equals<SubModelType, S['SubModelType']>>());
-  test('ClipVariantType', () => assert<Equals<z.infer<typeof zClipVariantType>, S['ClipVariantType']>>());
-  test('ModelVariantType', () => assert<Equals<z.infer<typeof zModelVariantType>, S['ModelVariantType']>>());
-  test('ModelFormat', () => assert<Equals<z.infer<typeof zModelFormat>, S['ModelFormat']>>());
+  test('ModelIdentifier', () => assert<Equals<ModelType, S['ModelType']>>());
 
   // Misc types
   test('ProgressImage', () => assert<Equals<ProgressImage, S['ProgressImage']>>());

invokeai/frontend/web/src/features/nodes/types/common.ts

@@ -73,7 +73,7 @@ export type SchedulerField = z.infer<typeof zSchedulerField>;
 // #endregion
 
 // #region Model-related schemas
-export const zBaseModelType = z.enum([
+const zBaseModel = z.enum([
   'any',
   'sd-1',
   'sd-2',
@@ -82,7 +82,6 @@ export const zBaseModelType = z.enum([
   'sdxl-refiner',
   'flux',
   'cogview4',
-  'qwen-image',
   'imagen3',
   'imagen4',
   'chatgpt-4o',
@@ -91,7 +90,7 @@ export const zBaseModelType = z.enum([
   'veo3',
   'runway',
 ]);
-export type BaseModelType = z.infer<typeof zBaseModelType>;
+export type BaseModelType = z.infer<typeof zBaseModel>;
 export const zMainModelBase = z.enum([
   'sd-1',
   'sd-2',
@@ -99,7 +98,6 @@ export const zMainModelBase = z.enum([
   'sdxl',
   'flux',
   'cogview4',
-  'qwen-image',
   'imagen3',
   'imagen4',
   'chatgpt-4o',
@@ -145,31 +143,11 @@ const zSubModelType = z.enum([
   'safety_checker',
 ]);
 export type SubModelType = z.infer<typeof zSubModelType>;
-
-export const zClipVariantType = z.enum(['large', 'gigantic']);
-export const zModelVariantType = z.enum(['normal', 'inpaint', 'depth']);
-export const zModelFormat = z.enum([
-  'omi',
-  'diffusers',
-  'checkpoint',
-  'lycoris',
-  'onnx',
-  'olive',
-  'embedding_file',
-  'embedding_folder',
-  'invokeai',
-  't5_encoder',
-  'bnb_quantized_int8b',
-  'bnb_quantized_nf4b',
-  'gguf_quantized',
-  'api',
-]);
-
 export const zModelIdentifierField = z.object({
   key: z.string().min(1),
   hash: z.string().min(1),
   name: z.string().min(1),
-  base: zBaseModelType,
+  base: zBaseModel,
   type: zModelType,
   submodel_type: zSubModelType.nullish(),
 });

invokeai/frontend/web/src/features/nodes/types/field.ts

@@ -9,18 +9,7 @@ import { assert } from 'tsafe';
 import { z } from 'zod';
 
 import type { ImageField } from './common';
-import {
-  zBaseModelType,
-  zBoardField,
-  zClipVariantType,
-  zColorField,
-  zImageField,
-  zModelFormat,
-  zModelIdentifierField,
-  zModelType,
-  zModelVariantType,
-  zSchedulerField,
-} from './common';
+import { zBoardField, zColorField, zImageField, zModelIdentifierField, zSchedulerField } from './common';
 
 /**
  * zod schemas & inferred types for fields.
@@ -71,10 +60,6 @@ const zFieldInputTemplateBase = zFieldTemplateBase.extend({
   default: z.undefined(),
   ui_component: zFieldUIComponent.nullish(),
   ui_choice_labels: z.record(z.string(), z.string()).nullish(),
-  ui_model_base: z.array(zBaseModelType).nullish(),
-  ui_model_type: z.array(zModelType).nullish(),
-  ui_model_variant: z.array(zModelVariantType.or(zClipVariantType)).nullish(),
-  ui_model_format: z.array(zModelFormat).nullish(),
 });
 const zFieldOutputTemplateBase = zFieldTemplateBase.extend({
   fieldKind: z.literal('output'),
@@ -176,10 +161,118 @@ const zColorFieldType = zFieldTypeBase.extend({
   name: z.literal('ColorField'),
   originalType: zStatelessFieldType.optional(),
 });
+const zMainModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('MainModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
 const zModelIdentifierFieldType = zFieldTypeBase.extend({
   name: z.literal('ModelIdentifierField'),
   originalType: zStatelessFieldType.optional(),
 });
+const zSDXLMainModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('SDXLMainModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zSD3MainModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('SD3MainModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zCogView4MainModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('CogView4MainModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zFluxMainModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('FluxMainModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zSDXLRefinerModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('SDXLRefinerModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zVAEModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('VAEModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zLoRAModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('LoRAModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zLLaVAModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('LLaVAModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zControlNetModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('ControlNetModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zIPAdapterModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('IPAdapterModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zT2IAdapterModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('T2IAdapterModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zSpandrelImageToImageModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('SpandrelImageToImageModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zT5EncoderModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('T5EncoderModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zCLIPEmbedModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('CLIPEmbedModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zCLIPLEmbedModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('CLIPLEmbedModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zCLIPGEmbedModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('CLIPGEmbedModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zControlLoRAModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('ControlLoRAModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zFluxVAEModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('FluxVAEModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zSigLipModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('SigLipModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zFluxReduxModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('FluxReduxModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zImagen3ModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('Imagen3ModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zImagen4ModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('Imagen4ModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zChatGPT4oModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('ChatGPT4oModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zVeo3ModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('Veo3ModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zRunwayModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('RunwayModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
+const zFluxKontextModelFieldType = zFieldTypeBase.extend({
+  name: z.literal('FluxKontextModelField'),
+  originalType: zStatelessFieldType.optional(),
+});
 const zSchedulerFieldType = zFieldTypeBase.extend({
   name: z.literal('SchedulerField'),
   originalType: zStatelessFieldType.optional(),
@@ -209,6 +302,33 @@ const zStatefulFieldType = z.union([
   zImageFieldType,
   zBoardFieldType,
   zModelIdentifierFieldType,
+  zMainModelFieldType,
+  zSDXLMainModelFieldType,
+  zSD3MainModelFieldType,
+  zCogView4MainModelFieldType,
+  zFluxMainModelFieldType,
+  zSDXLRefinerModelFieldType,
+  zVAEModelFieldType,
+  zLoRAModelFieldType,
+  zLLaVAModelFieldType,
+  zControlNetModelFieldType,
+  zIPAdapterModelFieldType,
+  zT2IAdapterModelFieldType,
+  zSpandrelImageToImageModelFieldType,
+  zT5EncoderModelFieldType,
+  zCLIPEmbedModelFieldType,
+  zCLIPLEmbedModelFieldType,
+  zCLIPGEmbedModelFieldType,
+  zControlLoRAModelFieldType,
+  zFluxVAEModelFieldType,
+  zSigLipModelFieldType,
+  zFluxReduxModelFieldType,
+  zImagen3ModelFieldType,
+  zImagen4ModelFieldType,
+  zChatGPT4oModelFieldType,
+  zFluxKontextModelFieldType,
+  zVeo3ModelFieldType,
+  zRunwayModelFieldType,
   zColorFieldType,
   zSchedulerFieldType,
   zFloatGeneratorFieldType,
@@ -224,7 +344,36 @@ const zFieldType = z.union([zStatefulFieldType, zStatelessFieldType]);
 export type FieldType = z.infer<typeof zFieldType>;
 
 const modelFieldTypeNames = [
+  // Stateful model fields
   zModelIdentifierFieldType.shape.name.value,
+  zMainModelFieldType.shape.name.value,
+  zSDXLMainModelFieldType.shape.name.value,
+  zSD3MainModelFieldType.shape.name.value,
+  zCogView4MainModelFieldType.shape.name.value,
+  zFluxMainModelFieldType.shape.name.value,
+  zSDXLRefinerModelFieldType.shape.name.value,
+  zVAEModelFieldType.shape.name.value,
+  zLoRAModelFieldType.shape.name.value,
+  zLLaVAModelFieldType.shape.name.value,
+  zControlNetModelFieldType.shape.name.value,
+  zIPAdapterModelFieldType.shape.name.value,
+  zT2IAdapterModelFieldType.shape.name.value,
+  zSpandrelImageToImageModelFieldType.shape.name.value,
+  zT5EncoderModelFieldType.shape.name.value,
+  zCLIPEmbedModelFieldType.shape.name.value,
+  zCLIPLEmbedModelFieldType.shape.name.value,
+  zCLIPGEmbedModelFieldType.shape.name.value,
+  zControlLoRAModelFieldType.shape.name.value,
+  zFluxVAEModelFieldType.shape.name.value,
+  zSigLipModelFieldType.shape.name.value,
+  zFluxReduxModelFieldType.shape.name.value,
+  zImagen3ModelFieldType.shape.name.value,
+  zImagen4ModelFieldType.shape.name.value,
+  zChatGPT4oModelFieldType.shape.name.value,
+  zFluxKontextModelFieldType.shape.name.value,
+  zVeo3ModelFieldType.shape.name.value,
+  zRunwayModelFieldType.shape.name.value,
+  // Stateless model fields
   'UNetField',
   'VAEField',
   'CLIPField',
@@ -630,6 +779,26 @@ export const isColorFieldInputInstance = buildInstanceTypeGuard(zColorFieldInput
 export const isColorFieldInputTemplate = buildTemplateTypeGuard<ColorFieldInputTemplate>('ColorField');
 // #endregion
 
+// #region MainModelField
+export const zMainModelFieldValue = zModelIdentifierField.optional();
+const zMainModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zMainModelFieldValue,
+});
+const zMainModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zMainModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zMainModelFieldValue,
+});
+const zMainModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zMainModelFieldType,
+});
+export type MainModelFieldValue = z.infer<typeof zMainModelFieldValue>;
+export type MainModelFieldInputInstance = z.infer<typeof zMainModelFieldInputInstance>;
+export type MainModelFieldInputTemplate = z.infer<typeof zMainModelFieldInputTemplate>;
+export const isMainModelFieldInputInstance = buildInstanceTypeGuard(zMainModelFieldInputInstance);
+export const isMainModelFieldInputTemplate = buildTemplateTypeGuard<MainModelFieldInputTemplate>('MainModelField');
+// #endregion
+
 // #region ModelIdentifierField
 export const zModelIdentifierFieldValue = zModelIdentifierField.optional();
 const zModelIdentifierFieldInputInstance = zFieldInputInstanceBase.extend({
@@ -651,6 +820,507 @@ export const isModelIdentifierFieldInputTemplate =
   buildTemplateTypeGuard<ModelIdentifierFieldInputTemplate>('ModelIdentifierField');
 // #endregion
 
+// #region SDXLMainModelField
+const zSDXLMainModelFieldValue = zMainModelFieldValue; // TODO: Narrow to SDXL models only.
+const zSDXLMainModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zSDXLMainModelFieldValue,
+});
+const zSDXLMainModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zSDXLMainModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zSDXLMainModelFieldValue,
+});
+const zSDXLMainModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zSDXLMainModelFieldType,
+});
+export type SDXLMainModelFieldInputInstance = z.infer<typeof zSDXLMainModelFieldInputInstance>;
+export type SDXLMainModelFieldInputTemplate = z.infer<typeof zSDXLMainModelFieldInputTemplate>;
+export const isSDXLMainModelFieldInputInstance = buildInstanceTypeGuard(zSDXLMainModelFieldInputInstance);
+export const isSDXLMainModelFieldInputTemplate =
+  buildTemplateTypeGuard<SDXLMainModelFieldInputTemplate>('SDXLMainModelField');
+// #endregion
+
+// #region SD3MainModelField
+const zSD3MainModelFieldValue = zMainModelFieldValue; // TODO: Narrow to SDXL models only.
+const zSD3MainModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zSD3MainModelFieldValue,
+});
+const zSD3MainModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zSD3MainModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zSD3MainModelFieldValue,
+});
+const zSD3MainModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zSD3MainModelFieldType,
+});
+export type SD3MainModelFieldInputInstance = z.infer<typeof zSD3MainModelFieldInputInstance>;
+export type SD3MainModelFieldInputTemplate = z.infer<typeof zSD3MainModelFieldInputTemplate>;
+export const isSD3MainModelFieldInputInstance = buildInstanceTypeGuard(zSD3MainModelFieldInputInstance);
+export const isSD3MainModelFieldInputTemplate =
+  buildTemplateTypeGuard<SD3MainModelFieldInputTemplate>('SD3MainModelField');
+// #endregion
+
+// #region CogView4MainModelField
+const zCogView4MainModelFieldValue = zMainModelFieldValue;
+const zCogView4MainModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zCogView4MainModelFieldValue,
+});
+const zCogView4MainModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zCogView4MainModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zCogView4MainModelFieldValue,
+});
+const zCogView4MainModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zCogView4MainModelFieldType,
+});
+export type CogView4MainModelFieldInputInstance = z.infer<typeof zCogView4MainModelFieldInputInstance>;
+export type CogView4MainModelFieldInputTemplate = z.infer<typeof zCogView4MainModelFieldInputTemplate>;
+export const isCogView4MainModelFieldInputInstance = buildInstanceTypeGuard(zCogView4MainModelFieldInputInstance);
+export const isCogView4MainModelFieldInputTemplate =
+  buildTemplateTypeGuard<CogView4MainModelFieldInputTemplate>('CogView4MainModelField');
+// #endregion
+
+// #region FluxMainModelField
+const zFluxMainModelFieldValue = zMainModelFieldValue; // TODO: Narrow to SDXL models only.
+const zFluxMainModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zFluxMainModelFieldValue,
+});
+const zFluxMainModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zFluxMainModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zFluxMainModelFieldValue,
+});
+const zFluxMainModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zFluxMainModelFieldType,
+});
+export type FluxMainModelFieldInputInstance = z.infer<typeof zFluxMainModelFieldInputInstance>;
+export type FluxMainModelFieldInputTemplate = z.infer<typeof zFluxMainModelFieldInputTemplate>;
+export const isFluxMainModelFieldInputInstance = buildInstanceTypeGuard(zFluxMainModelFieldInputInstance);
+export const isFluxMainModelFieldInputTemplate =
+  buildTemplateTypeGuard<FluxMainModelFieldInputTemplate>('FluxMainModelField');
+// #endregion
+
+// #region SDXLRefinerModelField
+/** @alias */ // tells knip to ignore this duplicate export
+export const zSDXLRefinerModelFieldValue = zMainModelFieldValue; // TODO: Narrow to SDXL Refiner models only.
+const zSDXLRefinerModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zSDXLRefinerModelFieldValue,
+});
+const zSDXLRefinerModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zSDXLRefinerModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zSDXLRefinerModelFieldValue,
+});
+const zSDXLRefinerModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zSDXLRefinerModelFieldType,
+});
+export type SDXLRefinerModelFieldValue = z.infer<typeof zSDXLRefinerModelFieldValue>;
+export type SDXLRefinerModelFieldInputInstance = z.infer<typeof zSDXLRefinerModelFieldInputInstance>;
+export type SDXLRefinerModelFieldInputTemplate = z.infer<typeof zSDXLRefinerModelFieldInputTemplate>;
+export const isSDXLRefinerModelFieldInputInstance = buildInstanceTypeGuard(zSDXLRefinerModelFieldInputInstance);
+export const isSDXLRefinerModelFieldInputTemplate =
+  buildTemplateTypeGuard<SDXLRefinerModelFieldInputTemplate>('SDXLRefinerModelField');
+// #endregion
+
+// #region VAEModelField
+
+export const zVAEModelFieldValue = zModelIdentifierField.optional();
+const zVAEModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zVAEModelFieldValue,
+});
+const zVAEModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zVAEModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zVAEModelFieldValue,
+});
+const zVAEModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zVAEModelFieldType,
+});
+export type VAEModelFieldValue = z.infer<typeof zVAEModelFieldValue>;
+export type VAEModelFieldInputInstance = z.infer<typeof zVAEModelFieldInputInstance>;
+export type VAEModelFieldInputTemplate = z.infer<typeof zVAEModelFieldInputTemplate>;
+export const isVAEModelFieldInputInstance = buildInstanceTypeGuard(zVAEModelFieldInputInstance);
+export const isVAEModelFieldInputTemplate = buildTemplateTypeGuard<VAEModelFieldInputTemplate>('VAEModelField');
+// #endregion
+
+// #region LoRAModelField
+export const zLoRAModelFieldValue = zModelIdentifierField.optional();
+const zLoRAModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zLoRAModelFieldValue,
+});
+const zLoRAModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zLoRAModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zLoRAModelFieldValue,
+});
+const zLoRAModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zLoRAModelFieldType,
+});
+export type LoRAModelFieldValue = z.infer<typeof zLoRAModelFieldValue>;
+export type LoRAModelFieldInputInstance = z.infer<typeof zLoRAModelFieldInputInstance>;
+export type LoRAModelFieldInputTemplate = z.infer<typeof zLoRAModelFieldInputTemplate>;
+export const isLoRAModelFieldInputInstance = buildInstanceTypeGuard(zLoRAModelFieldInputInstance);
+export const isLoRAModelFieldInputTemplate = buildTemplateTypeGuard<LoRAModelFieldInputTemplate>('LoRAModelField');
+// #endregion
+
+// #region LLaVAModelField
+export const zLLaVAModelFieldValue = zModelIdentifierField.optional();
+const zLLaVAModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zLLaVAModelFieldValue,
+});
+const zLLaVAModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zLLaVAModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zLLaVAModelFieldValue,
+});
+const zLLaVAModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zLLaVAModelFieldType,
+});
+export type LLaVAModelFieldValue = z.infer<typeof zLLaVAModelFieldValue>;
+export type LLaVAModelFieldInputInstance = z.infer<typeof zLLaVAModelFieldInputInstance>;
+export type LLaVAModelFieldInputTemplate = z.infer<typeof zLLaVAModelFieldInputTemplate>;
+export const isLLaVAModelFieldInputInstance = buildInstanceTypeGuard(zLLaVAModelFieldInputInstance);
+export const isLLaVAModelFieldInputTemplate = buildTemplateTypeGuard<LLaVAModelFieldInputTemplate>('LLaVAModelField');
+// #endregion
+
+// #region ControlNetModelField
+export const zControlNetModelFieldValue = zModelIdentifierField.optional();
+const zControlNetModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zControlNetModelFieldValue,
+});
+const zControlNetModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zControlNetModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zControlNetModelFieldValue,
+});
+const zControlNetModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zControlNetModelFieldType,
+});
+export type ControlNetModelFieldValue = z.infer<typeof zControlNetModelFieldValue>;
+export type ControlNetModelFieldInputInstance = z.infer<typeof zControlNetModelFieldInputInstance>;
+export type ControlNetModelFieldInputTemplate = z.infer<typeof zControlNetModelFieldInputTemplate>;
+export const isControlNetModelFieldInputInstance = buildInstanceTypeGuard(zControlNetModelFieldInputInstance);
+export const isControlNetModelFieldInputTemplate =
+  buildTemplateTypeGuard<ControlNetModelFieldInputTemplate>('ControlNetModelField');
+// #endregion
+
+// #region IPAdapterModelField
+export const zIPAdapterModelFieldValue = zModelIdentifierField.optional();
+const zIPAdapterModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zIPAdapterModelFieldValue,
+});
+const zIPAdapterModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zIPAdapterModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zIPAdapterModelFieldValue,
+});
+const zIPAdapterModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zIPAdapterModelFieldType,
+});
+export type IPAdapterModelFieldValue = z.infer<typeof zIPAdapterModelFieldValue>;
+export type IPAdapterModelFieldInputInstance = z.infer<typeof zIPAdapterModelFieldInputInstance>;
+export type IPAdapterModelFieldInputTemplate = z.infer<typeof zIPAdapterModelFieldInputTemplate>;
+export const isIPAdapterModelFieldInputInstance = buildInstanceTypeGuard(zIPAdapterModelFieldInputInstance);
+export const isIPAdapterModelFieldInputTemplate =
+  buildTemplateTypeGuard<IPAdapterModelFieldInputTemplate>('IPAdapterModelField');
+// #endregion
+
+// #region T2IAdapterField
+export const zT2IAdapterModelFieldValue = zModelIdentifierField.optional();
+const zT2IAdapterModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zT2IAdapterModelFieldValue,
+});
+const zT2IAdapterModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zT2IAdapterModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zT2IAdapterModelFieldValue,
+});
+const zT2IAdapterModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zT2IAdapterModelFieldType,
+});
+export type T2IAdapterModelFieldValue = z.infer<typeof zT2IAdapterModelFieldValue>;
+export type T2IAdapterModelFieldInputInstance = z.infer<typeof zT2IAdapterModelFieldInputInstance>;
+export type T2IAdapterModelFieldInputTemplate = z.infer<typeof zT2IAdapterModelFieldInputTemplate>;
+export const isT2IAdapterModelFieldInputInstance = buildInstanceTypeGuard(zT2IAdapterModelFieldInputInstance);
+export const isT2IAdapterModelFieldInputTemplate =
+  buildTemplateTypeGuard<T2IAdapterModelFieldInputTemplate>('T2IAdapterModelField');
+// #endregion
+
+// #region SpandrelModelToModelField
+export const zSpandrelImageToImageModelFieldValue = zModelIdentifierField.optional();
+const zSpandrelImageToImageModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zSpandrelImageToImageModelFieldValue,
+});
+const zSpandrelImageToImageModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zSpandrelImageToImageModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zSpandrelImageToImageModelFieldValue,
+});
+const zSpandrelImageToImageModelFieldOutputTemplate = zFieldOutputTemplateBase.extend({
+  type: zSpandrelImageToImageModelFieldType,
+});
+export type SpandrelImageToImageModelFieldValue = z.infer<typeof zSpandrelImageToImageModelFieldValue>;
+export type SpandrelImageToImageModelFieldInputInstance = z.infer<typeof zSpandrelImageToImageModelFieldInputInstance>;
+export type SpandrelImageToImageModelFieldInputTemplate = z.infer<typeof zSpandrelImageToImageModelFieldInputTemplate>;
+export const isSpandrelImageToImageModelFieldInputInstance = buildInstanceTypeGuard(
+  zSpandrelImageToImageModelFieldInputInstance
+);
+export const isSpandrelImageToImageModelFieldInputTemplate =
+  buildTemplateTypeGuard<SpandrelImageToImageModelFieldInputTemplate>('SpandrelImageToImageModelField');
+// #endregion
+
+// #region T5EncoderModelField
+
+export const zT5EncoderModelFieldValue = zModelIdentifierField.optional();
+const zT5EncoderModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zT5EncoderModelFieldValue,
+});
+const zT5EncoderModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zT5EncoderModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zT5EncoderModelFieldValue,
+});
+export type T5EncoderModelFieldValue = z.infer<typeof zT5EncoderModelFieldValue>;
+export type T5EncoderModelFieldInputInstance = z.infer<typeof zT5EncoderModelFieldInputInstance>;
+export type T5EncoderModelFieldInputTemplate = z.infer<typeof zT5EncoderModelFieldInputTemplate>;
+export const isT5EncoderModelFieldInputInstance = buildInstanceTypeGuard(zT5EncoderModelFieldInputInstance);
+export const isT5EncoderModelFieldInputTemplate =
+  buildTemplateTypeGuard<T5EncoderModelFieldInputTemplate>('T5EncoderModelField');
+// #endregion
+
+// #region FluxVAEModelField
+export const zFluxVAEModelFieldValue = zModelIdentifierField.optional();
+const zFluxVAEModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zFluxVAEModelFieldValue,
+});
+const zFluxVAEModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zFluxVAEModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zFluxVAEModelFieldValue,
+});
+export type FluxVAEModelFieldValue = z.infer<typeof zFluxVAEModelFieldValue>;
+export type FluxVAEModelFieldInputInstance = z.infer<typeof zFluxVAEModelFieldInputInstance>;
+export type FluxVAEModelFieldInputTemplate = z.infer<typeof zFluxVAEModelFieldInputTemplate>;
+export const isFluxVAEModelFieldInputInstance = buildInstanceTypeGuard(zFluxVAEModelFieldInputInstance);
+export const isFluxVAEModelFieldInputTemplate =
+  buildTemplateTypeGuard<FluxVAEModelFieldInputTemplate>('FluxVAEModelField');
+// #endregion
+
+// #region CLIPEmbedModelField
+export const zCLIPEmbedModelFieldValue = zModelIdentifierField.optional();
+const zCLIPEmbedModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zCLIPEmbedModelFieldValue,
+});
+const zCLIPEmbedModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zCLIPEmbedModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zCLIPEmbedModelFieldValue,
+});
+export type CLIPEmbedModelFieldValue = z.infer<typeof zCLIPEmbedModelFieldValue>;
+export type CLIPEmbedModelFieldInputInstance = z.infer<typeof zCLIPEmbedModelFieldInputInstance>;
+export type CLIPEmbedModelFieldInputTemplate = z.infer<typeof zCLIPEmbedModelFieldInputTemplate>;
+export const isCLIPEmbedModelFieldInputInstance = buildInstanceTypeGuard(zCLIPEmbedModelFieldInputInstance);
+export const isCLIPEmbedModelFieldInputTemplate =
+  buildTemplateTypeGuard<CLIPEmbedModelFieldInputTemplate>('CLIPEmbedModelField');
+// #endregion
+
+// #region CLIPLEmbedModelField
+export const zCLIPLEmbedModelFieldValue = zModelIdentifierField.optional();
+const zCLIPLEmbedModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zCLIPLEmbedModelFieldValue,
+});
+const zCLIPLEmbedModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zCLIPLEmbedModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zCLIPLEmbedModelFieldValue,
+});
+export type CLIPLEmbedModelFieldValue = z.infer<typeof zCLIPLEmbedModelFieldValue>;
+export type CLIPLEmbedModelFieldInputInstance = z.infer<typeof zCLIPLEmbedModelFieldInputInstance>;
+export type CLIPLEmbedModelFieldInputTemplate = z.infer<typeof zCLIPLEmbedModelFieldInputTemplate>;
+export const isCLIPLEmbedModelFieldInputInstance = buildInstanceTypeGuard(zCLIPLEmbedModelFieldInputInstance);
+export const isCLIPLEmbedModelFieldInputTemplate =
+  buildTemplateTypeGuard<CLIPLEmbedModelFieldInputTemplate>('CLIPLEmbedModelField');
+// #endregion
+
+// #region CLIPGEmbedModelField
+export const zCLIPGEmbedModelFieldValue = zModelIdentifierField.optional();
+const zCLIPGEmbedModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zCLIPGEmbedModelFieldValue,
+});
+const zCLIPGEmbedModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zCLIPGEmbedModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zCLIPGEmbedModelFieldValue,
+});
+export type CLIPGEmbedModelFieldValue = z.infer<typeof zCLIPLEmbedModelFieldValue>;
+export type CLIPGEmbedModelFieldInputInstance = z.infer<typeof zCLIPGEmbedModelFieldInputInstance>;
+export type CLIPGEmbedModelFieldInputTemplate = z.infer<typeof zCLIPGEmbedModelFieldInputTemplate>;
+export const isCLIPGEmbedModelFieldInputInstance = buildInstanceTypeGuard(zCLIPGEmbedModelFieldInputInstance);
+export const isCLIPGEmbedModelFieldInputTemplate =
+  buildTemplateTypeGuard<CLIPGEmbedModelFieldInputTemplate>('CLIPGEmbedModelField');
+// #endregion
+
+// #region ControlLoRAModelField
+export const zControlLoRAModelFieldValue = zModelIdentifierField.optional();
+const zControlLoRAModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zControlLoRAModelFieldValue,
+});
+const zControlLoRAModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zControlLoRAModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zControlLoRAModelFieldValue,
+});
+export type ControlLoRAModelFieldValue = z.infer<typeof zCLIPLEmbedModelFieldValue>;
+export type ControlLoRAModelFieldInputInstance = z.infer<typeof zControlLoRAModelFieldInputInstance>;
+export type ControlLoRAModelFieldInputTemplate = z.infer<typeof zControlLoRAModelFieldInputTemplate>;
+export const isControlLoRAModelFieldInputInstance = buildInstanceTypeGuard(zControlLoRAModelFieldInputInstance);
+export const isControlLoRAModelFieldInputTemplate =
+  buildTemplateTypeGuard<ControlLoRAModelFieldInputTemplate>('ControlLoRAModelField');
+// #endregion
+
+// #region SigLipModelField
+export const zSigLipModelFieldValue = zModelIdentifierField.optional();
+const zSigLipModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zSigLipModelFieldValue,
+});
+const zSigLipModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zSigLipModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zSigLipModelFieldValue,
+});
+export type SigLipModelFieldValue = z.infer<typeof zSigLipModelFieldValue>;
+export type SigLipModelFieldInputInstance = z.infer<typeof zSigLipModelFieldInputInstance>;
+export type SigLipModelFieldInputTemplate = z.infer<typeof zSigLipModelFieldInputTemplate>;
+export const isSigLipModelFieldInputInstance = buildInstanceTypeGuard(zSigLipModelFieldInputInstance);
+export const isSigLipModelFieldInputTemplate =
+  buildTemplateTypeGuard<SigLipModelFieldInputTemplate>('SigLipModelField');
+// #endregion
+
+// #region FluxReduxModelField
+export const zFluxReduxModelFieldValue = zModelIdentifierField.optional();
+const zFluxReduxModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zFluxReduxModelFieldValue,
+});
+const zFluxReduxModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zFluxReduxModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zFluxReduxModelFieldValue,
+});
+export type FluxReduxModelFieldValue = z.infer<typeof zFluxReduxModelFieldValue>;
+export type FluxReduxModelFieldInputInstance = z.infer<typeof zFluxReduxModelFieldInputInstance>;
+export type FluxReduxModelFieldInputTemplate = z.infer<typeof zFluxReduxModelFieldInputTemplate>;
+export const isFluxReduxModelFieldInputInstance = buildInstanceTypeGuard(zFluxReduxModelFieldInputInstance);
+export const isFluxReduxModelFieldInputTemplate =
+  buildTemplateTypeGuard<FluxReduxModelFieldInputTemplate>('FluxReduxModelField');
+// #endregion
+
+// #region Imagen3ModelField
+export const zImagen3ModelFieldValue = zModelIdentifierField.optional();
+const zImagen3ModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zImagen3ModelFieldValue,
+});
+const zImagen3ModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zImagen3ModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zImagen3ModelFieldValue,
+});
+export type Imagen3ModelFieldValue = z.infer<typeof zImagen3ModelFieldValue>;
+export type Imagen3ModelFieldInputInstance = z.infer<typeof zImagen3ModelFieldInputInstance>;
+export type Imagen3ModelFieldInputTemplate = z.infer<typeof zImagen3ModelFieldInputTemplate>;
+export const isImagen3ModelFieldInputInstance = buildInstanceTypeGuard(zImagen3ModelFieldInputInstance);
+export const isImagen3ModelFieldInputTemplate =
+  buildTemplateTypeGuard<Imagen3ModelFieldInputTemplate>('Imagen3ModelField');
+// #endregion
+
+// #region Imagen4ModelField
+export const zImagen4ModelFieldValue = zModelIdentifierField.optional();
+const zImagen4ModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zImagen4ModelFieldValue,
+});
+const zImagen4ModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zImagen4ModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zImagen4ModelFieldValue,
+});
+export type Imagen4ModelFieldValue = z.infer<typeof zImagen4ModelFieldValue>;
+export type Imagen4ModelFieldInputInstance = z.infer<typeof zImagen4ModelFieldInputInstance>;
+export type Imagen4ModelFieldInputTemplate = z.infer<typeof zImagen4ModelFieldInputTemplate>;
+export const isImagen4ModelFieldInputInstance = buildInstanceTypeGuard(zImagen4ModelFieldInputInstance);
+export const isImagen4ModelFieldInputTemplate =
+  buildTemplateTypeGuard<Imagen4ModelFieldInputTemplate>('Imagen4ModelField');
+// #endregion
+
+// #region FluxKontextModelField
+export const zFluxKontextModelFieldValue = zModelIdentifierField.optional();
+const zFluxKontextModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zFluxKontextModelFieldValue,
+});
+const zFluxKontextModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zFluxKontextModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zFluxKontextModelFieldValue,
+});
+export type FluxKontextModelFieldValue = z.infer<typeof zFluxKontextModelFieldValue>;
+export type FluxKontextModelFieldInputInstance = z.infer<typeof zFluxKontextModelFieldInputInstance>;
+export type FluxKontextModelFieldInputTemplate = z.infer<typeof zFluxKontextModelFieldInputTemplate>;
+export const isFluxKontextModelFieldInputInstance = buildInstanceTypeGuard(zFluxKontextModelFieldInputInstance);
+export const isFluxKontextModelFieldInputTemplate =
+  buildTemplateTypeGuard<FluxKontextModelFieldInputTemplate>('FluxKontextModelField');
+// #endregion
+
+// #region ChatGPT4oModelField
+export const zChatGPT4oModelFieldValue = zModelIdentifierField.optional();
+const zChatGPT4oModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zChatGPT4oModelFieldValue,
+});
+const zChatGPT4oModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zChatGPT4oModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zChatGPT4oModelFieldValue,
+});
+export type ChatGPT4oModelFieldValue = z.infer<typeof zChatGPT4oModelFieldValue>;
+export type ChatGPT4oModelFieldInputInstance = z.infer<typeof zChatGPT4oModelFieldInputInstance>;
+export type ChatGPT4oModelFieldInputTemplate = z.infer<typeof zChatGPT4oModelFieldInputTemplate>;
+export const isChatGPT4oModelFieldInputInstance = buildInstanceTypeGuard(zChatGPT4oModelFieldInputInstance);
+export const isChatGPT4oModelFieldInputTemplate =
+  buildTemplateTypeGuard<ChatGPT4oModelFieldInputTemplate>('ChatGPT4oModelField');
+// #endregion
+
+// #region Veo3ModelField
+export const zVeo3ModelFieldValue = zModelIdentifierField.optional();
+const zVeo3ModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zVeo3ModelFieldValue,
+});
+const zVeo3ModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zVeo3ModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zVeo3ModelFieldValue,
+});
+export type Veo3ModelFieldValue = z.infer<typeof zVeo3ModelFieldValue>;
+export type Veo3ModelFieldInputInstance = z.infer<typeof zVeo3ModelFieldInputInstance>;
+export type Veo3ModelFieldInputTemplate = z.infer<typeof zVeo3ModelFieldInputTemplate>;
+export const isVeo3ModelFieldInputInstance = buildInstanceTypeGuard(zVeo3ModelFieldInputInstance);
+export const isVeo3ModelFieldInputTemplate = buildTemplateTypeGuard<Veo3ModelFieldInputTemplate>('Veo3ModelField');
+// #endregion
+
+// #region RunwayModelField
+export const zRunwayModelFieldValue = zModelIdentifierField.optional();
+const zRunwayModelFieldInputInstance = zFieldInputInstanceBase.extend({
+  value: zRunwayModelFieldValue,
+});
+const zRunwayModelFieldInputTemplate = zFieldInputTemplateBase.extend({
+  type: zRunwayModelFieldType,
+  originalType: zFieldType.optional(),
+  default: zRunwayModelFieldValue,
+});
+export type RunwayModelFieldValue = z.infer<typeof zRunwayModelFieldValue>;
+export type RunwayModelFieldInputInstance = z.infer<typeof zRunwayModelFieldInputInstance>;
+export type RunwayModelFieldInputTemplate = z.infer<typeof zRunwayModelFieldInputTemplate>;
+export const isRunwayModelFieldInputInstance = buildInstanceTypeGuard(zRunwayModelFieldInputInstance);
+export const isRunwayModelFieldInputTemplate =
+  buildTemplateTypeGuard<RunwayModelFieldInputTemplate>('RunwayModelField');
+// #endregion
+
 // #region SchedulerField
 export const zSchedulerFieldValue = zSchedulerField.optional();
 const zSchedulerFieldInputInstance = zFieldInputInstanceBase.extend({
@@ -1261,6 +1931,31 @@ export const zStatefulFieldValue = z.union([
   zImageFieldCollectionValue,
   zBoardFieldValue,
   zModelIdentifierFieldValue,
+  zMainModelFieldValue,
+  zSDXLMainModelFieldValue,
+  zFluxMainModelFieldValue,
+  zSD3MainModelFieldValue,
+  zCogView4MainModelFieldValue,
+  zSDXLRefinerModelFieldValue,
+  zVAEModelFieldValue,
+  zLoRAModelFieldValue,
+  zLLaVAModelFieldValue,
+  zControlNetModelFieldValue,
+  zIPAdapterModelFieldValue,
+  zT2IAdapterModelFieldValue,
+  zSpandrelImageToImageModelFieldValue,
+  zT5EncoderModelFieldValue,
+  zFluxVAEModelFieldValue,
+  zCLIPEmbedModelFieldValue,
+  zCLIPLEmbedModelFieldValue,
+  zCLIPGEmbedModelFieldValue,
+  zControlLoRAModelFieldValue,
+  zSigLipModelFieldValue,
+  zFluxReduxModelFieldValue,
+  zImagen3ModelFieldValue,
+  zImagen4ModelFieldValue,
+  zFluxKontextModelFieldValue,
+  zChatGPT4oModelFieldValue,
   zColorFieldValue,
   zSchedulerFieldValue,
   zFloatGeneratorFieldValue,
@@ -1288,6 +1983,22 @@ const zStatefulFieldInputInstance = z.union([
   zImageFieldCollectionInputInstance,
   zBoardFieldInputInstance,
   zModelIdentifierFieldInputInstance,
+  zMainModelFieldInputInstance,
+  zFluxMainModelFieldInputInstance,
+  zSD3MainModelFieldInputInstance,
+  zCogView4MainModelFieldInputInstance,
+  zSDXLMainModelFieldInputInstance,
+  zSDXLRefinerModelFieldInputInstance,
+  zVAEModelFieldInputInstance,
+  zLoRAModelFieldInputInstance,
+  zLLaVAModelFieldInputInstance,
+  zControlNetModelFieldInputInstance,
+  zIPAdapterModelFieldInputInstance,
+  zT2IAdapterModelFieldInputInstance,
+  zSpandrelImageToImageModelFieldInputInstance,
+  zT5EncoderModelFieldInputInstance,
+  zFluxVAEModelFieldInputInstance,
+  zCLIPEmbedModelFieldInputInstance,
   zColorFieldInputInstance,
   zSchedulerFieldInputInstance,
   zFloatGeneratorFieldInputInstance,
@@ -1314,6 +2025,33 @@ const zStatefulFieldInputTemplate = z.union([
   zImageFieldCollectionInputTemplate,
   zBoardFieldInputTemplate,
   zModelIdentifierFieldInputTemplate,
+  zMainModelFieldInputTemplate,
+  zFluxMainModelFieldInputTemplate,
+  zSD3MainModelFieldInputTemplate,
+  zCogView4MainModelFieldInputTemplate,
+  zSDXLMainModelFieldInputTemplate,
+  zSDXLRefinerModelFieldInputTemplate,
+  zVAEModelFieldInputTemplate,
+  zLoRAModelFieldInputTemplate,
+  zLLaVAModelFieldInputTemplate,
+  zControlNetModelFieldInputTemplate,
+  zIPAdapterModelFieldInputTemplate,
+  zT2IAdapterModelFieldInputTemplate,
+  zSpandrelImageToImageModelFieldInputTemplate,
+  zT5EncoderModelFieldInputTemplate,
+  zFluxVAEModelFieldInputTemplate,
+  zCLIPEmbedModelFieldInputTemplate,
+  zCLIPLEmbedModelFieldInputTemplate,
+  zCLIPGEmbedModelFieldInputTemplate,
+  zControlLoRAModelFieldInputTemplate,
+  zSigLipModelFieldInputTemplate,
+  zFluxReduxModelFieldInputTemplate,
+  zImagen3ModelFieldInputTemplate,
+  zImagen4ModelFieldInputTemplate,
+  zChatGPT4oModelFieldInputTemplate,
+  zFluxKontextModelFieldInputTemplate,
+  zVeo3ModelFieldInputTemplate,
+  zRunwayModelFieldInputTemplate,
   zColorFieldInputTemplate,
   zSchedulerFieldInputTemplate,
   zStatelessFieldInputTemplate,
@@ -1341,6 +2079,19 @@ const zStatefulFieldOutputTemplate = z.union([
   zImageFieldCollectionOutputTemplate,
   zBoardFieldOutputTemplate,
   zModelIdentifierFieldOutputTemplate,
+  zMainModelFieldOutputTemplate,
+  zFluxMainModelFieldOutputTemplate,
+  zSD3MainModelFieldOutputTemplate,
+  zCogView4MainModelFieldOutputTemplate,
+  zSDXLMainModelFieldOutputTemplate,
+  zSDXLRefinerModelFieldOutputTemplate,
+  zVAEModelFieldOutputTemplate,
+  zLoRAModelFieldOutputTemplate,
+  zLLaVAModelFieldOutputTemplate,
+  zControlNetModelFieldOutputTemplate,
+  zIPAdapterModelFieldOutputTemplate,
+  zT2IAdapterModelFieldOutputTemplate,
+  zSpandrelImageToImageModelFieldOutputTemplate,
   zColorFieldOutputTemplate,
   zSchedulerFieldOutputTemplate,
   zFloatGeneratorFieldOutputTemplate,

invokeai/frontend/web/src/features/nodes/util/schema/buildFieldInputInstance.ts

@@ -9,9 +9,36 @@ const FIELD_VALUE_FALLBACK_MAP: Record<StatefulFieldType['name'], FieldValue> =
   FloatField: 0,
   ImageField: undefined,
   IntegerField: 0,
+  IPAdapterModelField: undefined,
+  LoRAModelField: undefined,
+  LLaVAModelField: undefined,
   ModelIdentifierField: undefined,
+  MainModelField: undefined,
   SchedulerField: 'dpmpp_3m_k',
+  SDXLMainModelField: undefined,
+  FluxMainModelField: undefined,
+  SD3MainModelField: undefined,
+  CogView4MainModelField: undefined,
+  SDXLRefinerModelField: undefined,
   StringField: '',
+  T2IAdapterModelField: undefined,
+  SpandrelImageToImageModelField: undefined,
+  VAEModelField: undefined,
+  ControlNetModelField: undefined,
+  T5EncoderModelField: undefined,
+  FluxVAEModelField: undefined,
+  CLIPEmbedModelField: undefined,
+  CLIPLEmbedModelField: undefined,
+  CLIPGEmbedModelField: undefined,
+  ControlLoRAModelField: undefined,
+  SigLipModelField: undefined,
+  FluxReduxModelField: undefined,
+  Imagen3ModelField: undefined,
+  Imagen4ModelField: undefined,
+  ChatGPT4oModelField: undefined,
+  FluxKontextModelField: undefined,
+  Veo3ModelField: undefined,
+  RunwayModelField: undefined,
   FloatGeneratorField: undefined,
   IntegerGeneratorField: undefined,
   StringGeneratorField: undefined,

invokeai/frontend/web/src/features/nodes/util/schema/buildFieldInputTemplate.ts

@@ -3,26 +3,53 @@ import { FieldParseError } from 'features/nodes/types/error';
 import type {
   BoardFieldInputTemplate,
   BooleanFieldInputTemplate,
+  ChatGPT4oModelFieldInputTemplate,
+  CLIPEmbedModelFieldInputTemplate,
+  CLIPGEmbedModelFieldInputTemplate,
+  CLIPLEmbedModelFieldInputTemplate,
+  CogView4MainModelFieldInputTemplate,
   ColorFieldInputTemplate,
+  ControlLoRAModelFieldInputTemplate,
+  ControlNetModelFieldInputTemplate,
   EnumFieldInputTemplate,
   FieldInputTemplate,
   FieldType,
   FloatFieldCollectionInputTemplate,
   FloatFieldInputTemplate,
   FloatGeneratorFieldInputTemplate,
+  FluxKontextModelFieldInputTemplate,
+  FluxMainModelFieldInputTemplate,
+  FluxReduxModelFieldInputTemplate,
+  FluxVAEModelFieldInputTemplate,
   ImageFieldCollectionInputTemplate,
   ImageFieldInputTemplate,
   ImageGeneratorFieldInputTemplate,
+  Imagen3ModelFieldInputTemplate,
+  Imagen4ModelFieldInputTemplate,
   IntegerFieldCollectionInputTemplate,
   IntegerFieldInputTemplate,
   IntegerGeneratorFieldInputTemplate,
+  IPAdapterModelFieldInputTemplate,
+  LLaVAModelFieldInputTemplate,
+  LoRAModelFieldInputTemplate,
+  MainModelFieldInputTemplate,
   ModelIdentifierFieldInputTemplate,
+  RunwayModelFieldInputTemplate,
   SchedulerFieldInputTemplate,
+  SD3MainModelFieldInputTemplate,
+  SDXLMainModelFieldInputTemplate,
+  SDXLRefinerModelFieldInputTemplate,
+  SigLipModelFieldInputTemplate,
+  SpandrelImageToImageModelFieldInputTemplate,
   StatefulFieldType,
   StatelessFieldInputTemplate,
   StringFieldCollectionInputTemplate,
   StringFieldInputTemplate,
   StringGeneratorFieldInputTemplate,
+  T2IAdapterModelFieldInputTemplate,
+  T5EncoderModelFieldInputTemplate,
+  VAEModelFieldInputTemplate,
+  Veo3ModelFieldInputTemplate,
 } from 'features/nodes/types/field';
 import {
   getFloatGeneratorArithmeticSequenceDefaults,
@@ -275,6 +302,373 @@ const buildModelIdentifierFieldInputTemplate: FieldInputTemplateBuilder<ModelIde
   return template;
 };
 
+const buildMainModelFieldInputTemplate: FieldInputTemplateBuilder<MainModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: MainModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildSDXLMainModelFieldInputTemplate: FieldInputTemplateBuilder<SDXLMainModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: SDXLMainModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildFluxMainModelFieldInputTemplate: FieldInputTemplateBuilder<FluxMainModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: FluxMainModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildSD3MainModelFieldInputTemplate: FieldInputTemplateBuilder<SD3MainModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: SD3MainModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildCogView4MainModelFieldInputTemplate: FieldInputTemplateBuilder<CogView4MainModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: CogView4MainModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildRefinerModelFieldInputTemplate: FieldInputTemplateBuilder<SDXLRefinerModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: SDXLRefinerModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildVAEModelFieldInputTemplate: FieldInputTemplateBuilder<VAEModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: VAEModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildT5EncoderModelFieldInputTemplate: FieldInputTemplateBuilder<T5EncoderModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: T5EncoderModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildCLIPEmbedModelFieldInputTemplate: FieldInputTemplateBuilder<CLIPEmbedModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: CLIPEmbedModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildCLIPLEmbedModelFieldInputTemplate: FieldInputTemplateBuilder<CLIPLEmbedModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: CLIPLEmbedModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildCLIPGEmbedModelFieldInputTemplate: FieldInputTemplateBuilder<CLIPGEmbedModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: CLIPGEmbedModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildControlLoRAModelFieldInputTemplate: FieldInputTemplateBuilder<ControlLoRAModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: ControlLoRAModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildLLaVAModelFieldInputTemplate: FieldInputTemplateBuilder<LLaVAModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: LLaVAModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildFluxVAEModelFieldInputTemplate: FieldInputTemplateBuilder<FluxVAEModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: FluxVAEModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildLoRAModelFieldInputTemplate: FieldInputTemplateBuilder<LoRAModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: LoRAModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildControlNetModelFieldInputTemplate: FieldInputTemplateBuilder<ControlNetModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: ControlNetModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildIPAdapterModelFieldInputTemplate: FieldInputTemplateBuilder<IPAdapterModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: IPAdapterModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildT2IAdapterModelFieldInputTemplate: FieldInputTemplateBuilder<T2IAdapterModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: T2IAdapterModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildSpandrelImageToImageModelFieldInputTemplate: FieldInputTemplateBuilder<
+  SpandrelImageToImageModelFieldInputTemplate
+> = ({ schemaObject, baseField, fieldType }) => {
+  const template: SpandrelImageToImageModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+
+  return template;
+};
+
+const buildSigLipModelFieldInputTemplate: FieldInputTemplateBuilder<SigLipModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: SigLipModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildFluxReduxModelFieldInputTemplate: FieldInputTemplateBuilder<FluxReduxModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: FluxReduxModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildImagen3ModelFieldInputTemplate: FieldInputTemplateBuilder<Imagen3ModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: Imagen3ModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildImagen4ModelFieldInputTemplate: FieldInputTemplateBuilder<Imagen4ModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: Imagen4ModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildFluxKontextModelFieldInputTemplate: FieldInputTemplateBuilder<FluxKontextModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: FluxKontextModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildVeo3ModelFieldInputTemplate: FieldInputTemplateBuilder<Veo3ModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: Veo3ModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildRunwayModelFieldInputTemplate: FieldInputTemplateBuilder<RunwayModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: RunwayModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
+const buildChatGPT4oModelFieldInputTemplate: FieldInputTemplateBuilder<ChatGPT4oModelFieldInputTemplate> = ({
+  schemaObject,
+  baseField,
+  fieldType,
+}) => {
+  const template: ChatGPT4oModelFieldInputTemplate = {
+    ...baseField,
+    type: fieldType,
+    default: schemaObject.default ?? undefined,
+  };
+  return template;
+};
+
 const buildBoardFieldInputTemplate: FieldInputTemplateBuilder<BoardFieldInputTemplate> = ({
   schemaObject,
   baseField,
@@ -449,41 +843,56 @@ const buildImageGeneratorFieldInputTemplate: FieldInputTemplateBuilder<ImageGene
   return template;
 };
 
-const TEMPLATE_BUILDER_MAP: Record<StatefulFieldType['name'], FieldInputTemplateBuilder> = {
+export const TEMPLATE_BUILDER_MAP: Record<StatefulFieldType['name'], FieldInputTemplateBuilder> = {
   BoardField: buildBoardFieldInputTemplate,
   BooleanField: buildBooleanFieldInputTemplate,
   ColorField: buildColorFieldInputTemplate,
+  ControlNetModelField: buildControlNetModelFieldInputTemplate,
   EnumField: buildEnumFieldInputTemplate,
   FloatField: buildFloatFieldInputTemplate,
   ImageField: buildImageFieldInputTemplate,
   IntegerField: buildIntegerFieldInputTemplate,
+  IPAdapterModelField: buildIPAdapterModelFieldInputTemplate,
+  LoRAModelField: buildLoRAModelFieldInputTemplate,
+  LLaVAModelField: buildLLaVAModelFieldInputTemplate,
   ModelIdentifierField: buildModelIdentifierFieldInputTemplate,
+  MainModelField: buildMainModelFieldInputTemplate,
   SchedulerField: buildSchedulerFieldInputTemplate,
+  SDXLMainModelField: buildSDXLMainModelFieldInputTemplate,
+  SD3MainModelField: buildSD3MainModelFieldInputTemplate,
+  CogView4MainModelField: buildCogView4MainModelFieldInputTemplate,
+  FluxMainModelField: buildFluxMainModelFieldInputTemplate,
+  SDXLRefinerModelField: buildRefinerModelFieldInputTemplate,
   StringField: buildStringFieldInputTemplate,
+  T2IAdapterModelField: buildT2IAdapterModelFieldInputTemplate,
+  SpandrelImageToImageModelField: buildSpandrelImageToImageModelFieldInputTemplate,
+  VAEModelField: buildVAEModelFieldInputTemplate,
+  T5EncoderModelField: buildT5EncoderModelFieldInputTemplate,
+  CLIPEmbedModelField: buildCLIPEmbedModelFieldInputTemplate,
+  CLIPLEmbedModelField: buildCLIPLEmbedModelFieldInputTemplate,
+  CLIPGEmbedModelField: buildCLIPGEmbedModelFieldInputTemplate,
+  FluxVAEModelField: buildFluxVAEModelFieldInputTemplate,
+  ControlLoRAModelField: buildControlLoRAModelFieldInputTemplate,
+  SigLipModelField: buildSigLipModelFieldInputTemplate,
+  FluxReduxModelField: buildFluxReduxModelFieldInputTemplate,
+  Imagen3ModelField: buildImagen3ModelFieldInputTemplate,
+  Imagen4ModelField: buildImagen4ModelFieldInputTemplate,
+  ChatGPT4oModelField: buildChatGPT4oModelFieldInputTemplate,
+  FluxKontextModelField: buildFluxKontextModelFieldInputTemplate,
+  Veo3ModelField: buildVeo3ModelFieldInputTemplate,
+  RunwayModelField: buildRunwayModelFieldInputTemplate,
   FloatGeneratorField: buildFloatGeneratorFieldInputTemplate,
   IntegerGeneratorField: buildIntegerGeneratorFieldInputTemplate,
   StringGeneratorField: buildStringGeneratorFieldInputTemplate,
   ImageGeneratorField: buildImageGeneratorFieldInputTemplate,
-};
+} as const;
 
 export const buildFieldInputTemplate = (
   fieldSchema: InvocationFieldSchema,
   fieldName: string,
   fieldType: FieldType
 ): FieldInputTemplate => {
-  const {
-    input,
-    ui_hidden,
-    ui_component,
-    ui_type,
-    ui_order,
-    ui_choice_labels,
-    orig_required: required,
-    ui_model_base,
-    ui_model_type,
-    ui_model_variant,
-    ui_model_format,
-  } = fieldSchema;
+  const { input, ui_hidden, ui_component, ui_type, ui_order, ui_choice_labels, orig_required: required } = fieldSchema;
 
   // This is the base field template that is common to all fields. The builder function will add all other
   // properties to this template.
@@ -499,10 +908,6 @@ export const buildFieldInputTemplate = (
     ui_type,
     ui_order,
     ui_choice_labels,
-    ui_model_base,
-    ui_model_type,
-    ui_model_variant,
-    ui_model_format,
   };
 
   if (isStatefulFieldType(fieldType)) {

invokeai/frontend/web/src/features/parameters/types/constants.ts

@@ -13,7 +13,6 @@ export const MODEL_TYPE_MAP: Record<BaseModelType, string> = {
   'sdxl-refiner': 'Stable Diffusion XL Refiner',
   flux: 'FLUX',
   cogview4: 'CogView4',
-  'qwen-image': 'Qwen-Image',
   imagen3: 'Imagen3',
   imagen4: 'Imagen4',
   'chatgpt-4o': 'ChatGPT 4o',
@@ -35,7 +34,6 @@ export const MODEL_TYPE_SHORT_MAP: Record<BaseModelType, string> = {
   'sdxl-refiner': 'SDXLR',
   flux: 'FLUX',
   cogview4: 'CogView4',
-  'qwen-image': 'Qwen',
   imagen3: 'Imagen3',
   imagen4: 'Imagen4',
   'chatgpt-4o': 'ChatGPT 4o',

invokeai/frontend/web/src/services/api/hooks/modelsByType.ts

@@ -12,25 +12,34 @@ import {
   isChatGPT4oModelConfig,
   isCLIPEmbedModelConfig,
   isCLIPVisionModelConfig,
+  isCogView4MainModelModelConfig,
   isControlLayerModelConfig,
   isControlLoRAModelConfig,
   isControlNetModelConfig,
   isFluxKontextApiModelConfig,
   isFluxKontextModelConfig,
+  isFluxMainModelModelConfig,
   isFluxReduxModelConfig,
   isFluxVAEModelConfig,
   isGemini2_5ModelConfig,
+  isImagen3ModelConfig,
+  isImagen4ModelConfig,
   isIPAdapterModelConfig,
   isLLaVAModelConfig,
   isLoRAModelConfig,
   isNonRefinerMainModelConfig,
+  isNonSDXLMainModelConfig,
   isRefinerMainModelModelConfig,
+  isRunwayModelConfig,
+  isSD3MainModelModelConfig,
+  isSDXLMainModelModelConfig,
   isSigLipModelConfig,
   isSpandrelImageToImageModelConfig,
   isT2IAdapterModelConfig,
   isT5EncoderModelConfig,
   isTIModelConfig,
   isVAEModelConfig,
+  isVeo3ModelConfig,
   isVideoModelConfig,
 } from 'services/api/types';
 
@@ -57,7 +66,12 @@ const buildModelsHook =
     return [modelConfigs, result] as const;
   };
 export const useMainModels = buildModelsHook(isNonRefinerMainModelConfig);
+export const useNonSDXLMainModels = buildModelsHook(isNonSDXLMainModelConfig);
 export const useRefinerModels = buildModelsHook(isRefinerMainModelModelConfig);
+export const useFluxModels = buildModelsHook(isFluxMainModelModelConfig);
+export const useSD3Models = buildModelsHook(isSD3MainModelModelConfig);
+export const useCogView4Models = buildModelsHook(isCogView4MainModelModelConfig);
+export const useSDXLModels = buildModelsHook(isSDXLMainModelModelConfig);
 export const useLoRAModels = buildModelsHook(isLoRAModelConfig);
 export const useControlLoRAModel = buildModelsHook(isControlLoRAModelConfig);
 export const useControlLayerModels = buildModelsHook(isControlLayerModelConfig);
@@ -89,6 +103,12 @@ export const useRegionalReferenceImageModels = buildModelsHook(
   (config) => isIPAdapterModelConfig(config) || isFluxReduxModelConfig(config)
 );
 export const useLLaVAModels = buildModelsHook(isLLaVAModelConfig);
+export const useImagen3Models = buildModelsHook(isImagen3ModelConfig);
+export const useImagen4Models = buildModelsHook(isImagen4ModelConfig);
+export const useChatGPT4oModels = buildModelsHook(isChatGPT4oModelConfig);
+export const useFluxKontextModels = buildModelsHook(isFluxKontextApiModelConfig);
+export const useVeo3Models = buildModelsHook(isVeo3ModelConfig);
+export const useRunwayModels = buildModelsHook(isRunwayModelConfig);
 export const useVideoModels = buildModelsHook(isVideoModelConfig);
 
 const buildModelsSelector =

invokeai/frontend/web/src/services/api/types.ts

@@ -122,7 +122,7 @@ export type T2IAdapterModelConfig = S['T2IAdapterConfig'];
 export type CLIPLEmbedModelConfig = S['CLIPLEmbedDiffusersConfig'];
 export type CLIPGEmbedModelConfig = S['CLIPGEmbedDiffusersConfig'];
 export type CLIPEmbedModelConfig = CLIPLEmbedModelConfig | CLIPGEmbedModelConfig;
-type LlavaOnevisionConfig = S['LlavaOnevisionConfig'];
+export type LlavaOnevisionConfig = S['LlavaOnevisionConfig'];
 export type T5EncoderModelConfig = S['T5EncoderConfig'];
 export type T5EncoderBnbQuantizedLlmInt8bModelConfig = S['T5EncoderBnbQuantizedLlmInt8bConfig'];
 export type SpandrelImageToImageModelConfig = S['SpandrelImageToImageConfig'];
@@ -130,14 +130,16 @@ type TextualInversionModelConfig = S['TextualInversionFileConfig'] | S['TextualI
 type DiffusersModelConfig = S['MainDiffusersConfig'];
 export type CheckpointModelConfig = S['MainCheckpointConfig'];
 type CLIPVisionDiffusersConfig = S['CLIPVisionDiffusersConfig'];
-type SigLipModelConfig = S['SigLIPConfig'];
+export type SigLipModelConfig = S['SigLIPConfig'];
 export type FLUXReduxModelConfig = S['FluxReduxConfig'];
-type ApiModelConfig = S['ApiModelConfig'];
+export type ApiModelConfig = S['ApiModelConfig'];
 export type VideoApiModelConfig = S['VideoApiModelConfig'];
 export type MainModelConfig = DiffusersModelConfig | CheckpointModelConfig | ApiModelConfig;
 export type FLUXKontextModelConfig = MainModelConfig;
 export type ChatGPT4oModelConfig = ApiModelConfig;
 export type Gemini2_5ModelConfig = ApiModelConfig;
+type Veo3ModelConfig = VideoApiModelConfig;
+type RunwayModelConfig = VideoApiModelConfig;
 export type AnyModelConfig =
   | ControlLoRAModelConfig
   | LoRAModelConfig
@@ -308,6 +310,22 @@ export const isVideoModelConfig = (config: AnyModelConfig): config is VideoApiMo
   return config.type === 'video';
 };
 
+export const isVeo3ModelConfig = (config: AnyModelConfig): config is Veo3ModelConfig => {
+  return config.base === 'veo3' && config.type === 'video';
+};
+
+export const isRunwayModelConfig = (config: AnyModelConfig): config is RunwayModelConfig => {
+  return config.base === 'runway' && config.type === 'video';
+};
+
+export const isImagen3ModelConfig = (config: AnyModelConfig): config is ApiModelConfig => {
+  return config.type === 'main' && config.base === 'imagen3';
+};
+
+export const isImagen4ModelConfig = (config: AnyModelConfig): config is ApiModelConfig => {
+  return config.type === 'main' && config.base === 'imagen4';
+};
+
 export const isFluxKontextApiModelConfig = (config: AnyModelConfig): config is ApiModelConfig => {
   return config.type === 'main' && config.base === 'flux-kontext';
 };
@@ -332,10 +350,30 @@ export const isRefinerMainModelModelConfig = (config: AnyModelConfig): config is
   return config.type === 'main' && config.base === 'sdxl-refiner';
 };
 
+export const isSDXLMainModelModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
+  return config.type === 'main' && config.base === 'sdxl';
+};
+
+export const isSD3MainModelModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
+  return config.type === 'main' && config.base === 'sd-3';
+};
+
+export const isCogView4MainModelModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
+  return config.type === 'main' && config.base === 'cogview4';
+};
+
+export const isFluxMainModelModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
+  return config.type === 'main' && config.base === 'flux';
+};
+
 export const isFluxFillMainModelModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
   return config.type === 'main' && config.base === 'flux' && config.variant === 'inpaint';
 };
 
+export const isNonSDXLMainModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
+  return config.type === 'main' && (config.base === 'sd-1' || config.base === 'sd-2');
+};
+
 export const isTIModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
   return config.type === 'embedding';
 };

pyproject.toml

@@ -36,7 +36,7 @@ dependencies = [
   "accelerate",
   "bitsandbytes; sys_platform!='darwin'",
   "compel==2.1.1",
-  "diffusers[torch]==0.35.0",
+  "diffusers[torch]==0.33.0",
   "gguf",
   "mediapipe==0.10.14",                   # needed for "mediapipeface" controlnet model
   "numpy<2.0.0",

qwen_test_config.yaml

@@ -1,26 +0,0 @@
-# Qwen-Image Test Configuration with Memory Optimizations
-# This config helps test Qwen-Image with limited VRAM
-
-# Model Cache Settings - Adjust based on your system
-# These settings enable CPU offloading for large models
-Model:
-  # Reduce VRAM cache to force CPU offloading
-  vram_cache_size: 8.0  # GB - Keep only essential models in VRAM
-  
-  # Increase RAM cache for CPU offloading
-  ram_cache_size: 32.0  # GB - Adjust based on available system RAM
-  
-  # Enable sequential offloading
-  sequential_offload: true
-  
-  # Use bfloat16 by default for all models
-  precision: bfloat16
-
-# Recommended workflow for testing:
-# 1. Load only the Qwen-Image model first (not Qwen2.5-VL)
-# 2. Use a simple text prompt without the text encoder
-# 3. Test with smaller image sizes (512x512) initially
-
-# Alternative: Use quantized models
-# Download: huggingface-cli download diffusers/qwen-image-nf4
-# This reduces memory usage by ~75%

run_qwen_optimized.sh

@@ -1,26 +0,0 @@
-#!/bin/bash
-# Run InvokeAI with optimized settings for Qwen-Image models
-
-echo "Starting InvokeAI with Qwen-Image memory optimizations..."
-echo "----------------------------------------"
-echo "Recommendations for 24GB VRAM systems:"
-echo "1. Set VRAM cache to 8-10GB in InvokeAI settings"
-echo "2. Set RAM cache to 20-30GB (based on available system RAM)"
-echo "3. Use bfloat16 precision (default in our loader)"
-echo "----------------------------------------"
-
-# Set environment variables for better memory management
-export PYTORCH_CUDA_ALLOC_CONF="max_split_size_mb:512"
-export CUDA_LAUNCH_BLOCKING=0
-
-# Optional: Limit CPU threads to prevent memory thrashing
-export OMP_NUM_THREADS=8
-
-# Run InvokeAI with your root directory
-invokeai-web --root ~/invokeai/ \
-  --precision bfloat16 \
-  --max_cache_size 8.0 \
-  --max_vram_cache_size 8.0
-
-# Alternative: Use with config file
-# invokeai-web --root ~/invokeai/ --config qwen_test_config.yaml