diff --git a/autogpt_platform/backend/backend/blocks/llm.py b/autogpt_platform/backend/backend/blocks/llm.py index 74f53d4bf1..30b9db770b 100644 --- a/autogpt_platform/backend/backend/blocks/llm.py +++ b/autogpt_platform/backend/backend/blocks/llm.py @@ -118,13 +118,8 @@ class LlmModel(str, Enum, metaclass=LlmModelMeta): AIML_API_META_LLAMA_3_1_70B = "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" AIML_API_LLAMA_3_2_3B = "meta-llama/Llama-3.2-3B-Instruct-Turbo" # Groq models - GEMMA2_9B = "gemma2-9b-it" LLAMA3_3_70B = "llama-3.3-70b-versatile" LLAMA3_1_8B = "llama-3.1-8b-instant" - LLAMA3_70B = "llama3-70b-8192" - LLAMA3_8B = "llama3-8b-8192" - # Groq preview models - DEEPSEEK_LLAMA_70B = "deepseek-r1-distill-llama-70b" # Ollama models OLLAMA_LLAMA3_3 = "llama3.3" OLLAMA_LLAMA3_2 = "llama3.2" @@ -134,7 +129,6 @@ class LlmModel(str, Enum, metaclass=LlmModelMeta): # OpenRouter models OPENAI_GPT_OSS_120B = "openai/gpt-oss-120b" OPENAI_GPT_OSS_20B = "openai/gpt-oss-20b" - GEMINI_FLASH_1_5 = "google/gemini-flash-1.5" GEMINI_2_5_PRO = "google/gemini-2.5-pro-preview-03-25" GEMINI_2_5_FLASH = "google/gemini-2.5-flash" GEMINI_2_0_FLASH = "google/gemini-2.0-flash-001" @@ -238,12 +232,8 @@ MODEL_METADATA = { LlmModel.AIML_API_META_LLAMA_3_1_70B: ModelMetadata("aiml_api", 131000, 2000), LlmModel.AIML_API_LLAMA_3_2_3B: ModelMetadata("aiml_api", 128000, None), # https://console.groq.com/docs/models - LlmModel.GEMMA2_9B: ModelMetadata("groq", 8192, None), LlmModel.LLAMA3_3_70B: ModelMetadata("groq", 128000, 32768), LlmModel.LLAMA3_1_8B: ModelMetadata("groq", 128000, 8192), - LlmModel.LLAMA3_70B: ModelMetadata("groq", 8192, None), - LlmModel.LLAMA3_8B: ModelMetadata("groq", 8192, None), - LlmModel.DEEPSEEK_LLAMA_70B: ModelMetadata("groq", 128000, None), # https://ollama.com/library LlmModel.OLLAMA_LLAMA3_3: ModelMetadata("ollama", 8192, None), LlmModel.OLLAMA_LLAMA3_2: ModelMetadata("ollama", 8192, None), @@ -251,7 +241,6 @@ MODEL_METADATA = { LlmModel.OLLAMA_LLAMA3_405B: ModelMetadata("ollama", 8192, None), LlmModel.OLLAMA_DOLPHIN: ModelMetadata("ollama", 32768, None), # https://openrouter.ai/models - LlmModel.GEMINI_FLASH_1_5: ModelMetadata("open_router", 1000000, 8192), LlmModel.GEMINI_2_5_PRO: ModelMetadata("open_router", 1050000, 8192), LlmModel.GEMINI_2_5_FLASH: ModelMetadata("open_router", 1048576, 65535), LlmModel.GEMINI_2_0_FLASH: ModelMetadata("open_router", 1048576, 8192), diff --git a/autogpt_platform/backend/backend/data/block_cost_config.py b/autogpt_platform/backend/backend/data/block_cost_config.py index d415673c84..6a54dfcc4d 100644 --- a/autogpt_platform/backend/backend/data/block_cost_config.py +++ b/autogpt_platform/backend/backend/data/block_cost_config.py @@ -82,20 +82,15 @@ MODEL_COST: dict[LlmModel, int] = { LlmModel.AIML_API_LLAMA3_3_70B: 1, LlmModel.AIML_API_META_LLAMA_3_1_70B: 1, LlmModel.AIML_API_LLAMA_3_2_3B: 1, - LlmModel.LLAMA3_8B: 1, - LlmModel.LLAMA3_70B: 1, - LlmModel.GEMMA2_9B: 1, LlmModel.LLAMA3_3_70B: 1, # $0.59 / $0.79 LlmModel.LLAMA3_1_8B: 1, LlmModel.OLLAMA_LLAMA3_3: 1, LlmModel.OLLAMA_LLAMA3_2: 1, LlmModel.OLLAMA_LLAMA3_8B: 1, LlmModel.OLLAMA_LLAMA3_405B: 1, - LlmModel.DEEPSEEK_LLAMA_70B: 1, # ? / ? LlmModel.OLLAMA_DOLPHIN: 1, LlmModel.OPENAI_GPT_OSS_120B: 1, LlmModel.OPENAI_GPT_OSS_20B: 1, - LlmModel.GEMINI_FLASH_1_5: 1, LlmModel.GEMINI_2_5_PRO: 4, LlmModel.MISTRAL_NEMO: 1, LlmModel.COHERE_COMMAND_R_08_2024: 1, diff --git a/autogpt_platform/backend/migrations/20251106091413_migrate_deprecated_groq_openrouter_models/migration.sql b/autogpt_platform/backend/migrations/20251106091413_migrate_deprecated_groq_openrouter_models/migration.sql new file mode 100644 index 0000000000..c750301ae0 --- /dev/null +++ b/autogpt_platform/backend/migrations/20251106091413_migrate_deprecated_groq_openrouter_models/migration.sql @@ -0,0 +1,53 @@ +-- Migrate deprecated Groq and OpenRouter models to their replacements +-- This updates all AgentNode blocks that use deprecated models that have been decommissioned +-- Deprecated models: +-- - deepseek-r1-distill-llama-70b (Groq - decommissioned) +-- - gemma2-9b-it (Groq - decommissioned) +-- - llama3-70b-8192 (Groq - decommissioned) +-- - llama3-8b-8192 (Groq - decommissioned) +-- - google/gemini-flash-1.5 (OpenRouter - no endpoints found) + +-- Update llama3-70b-8192 to llama-3.3-70b-versatile +UPDATE "AgentNode" +SET "constantInput" = JSONB_SET( + "constantInput"::jsonb, + '{model}', + '"llama-3.3-70b-versatile"'::jsonb + ) +WHERE "constantInput"::jsonb->>'model' = 'llama3-70b-8192'; + +-- Update llama3-8b-8192 to llama-3.1-8b-instant +UPDATE "AgentNode" +SET "constantInput" = JSONB_SET( + "constantInput"::jsonb, + '{model}', + '"llama-3.1-8b-instant"'::jsonb + ) +WHERE "constantInput"::jsonb->>'model' = 'llama3-8b-8192'; + +-- Update google/gemini-flash-1.5 to google/gemini-2.5-flash +UPDATE "AgentNode" +SET "constantInput" = JSONB_SET( + "constantInput"::jsonb, + '{model}', + '"google/gemini-2.5-flash"'::jsonb + ) +WHERE "constantInput"::jsonb->>'model' = 'google/gemini-flash-1.5'; + +-- Update deepseek-r1-distill-llama-70b to gpt-5-chat-latest (no direct replacement) +UPDATE "AgentNode" +SET "constantInput" = JSONB_SET( + "constantInput"::jsonb, + '{model}', + '"gpt-5-chat-latest"'::jsonb + ) +WHERE "constantInput"::jsonb->>'model' = 'deepseek-r1-distill-llama-70b'; + +-- Update gemma2-9b-it to gpt-5-chat-latest (no direct replacement) +UPDATE "AgentNode" +SET "constantInput" = JSONB_SET( + "constantInput"::jsonb, + '{model}', + '"gpt-5-chat-latest"'::jsonb + ) +WHERE "constantInput"::jsonb->>'model' = 'gemma2-9b-it'; diff --git a/classic/forge/forge/llm/providers/groq.py b/classic/forge/forge/llm/providers/groq.py index dc9e77e0b8..3b1dcacfd9 100644 --- a/classic/forge/forge/llm/providers/groq.py +++ b/classic/forge/forge/llm/providers/groq.py @@ -22,8 +22,6 @@ from .schema import ( class GroqModelName(str, enum.Enum): - LLAMA3_8B = "llama3-8b-8192" - LLAMA3_70B = "llama3-70b-8192" MIXTRAL_8X7B = "mixtral-8x7b-32768" GEMMA_7B = "gemma-7b-it" @@ -31,22 +29,6 @@ class GroqModelName(str, enum.Enum): GROQ_CHAT_MODELS = { info.name: info for info in [ - ChatModelInfo( - name=GroqModelName.LLAMA3_8B, - provider_name=ModelProviderName.GROQ, - prompt_token_cost=0.05 / 1e6, - completion_token_cost=0.10 / 1e6, - max_tokens=8192, - has_function_call_api=True, - ), - ChatModelInfo( - name=GroqModelName.LLAMA3_70B, - provider_name=ModelProviderName.GROQ, - prompt_token_cost=0.59 / 1e6, - completion_token_cost=0.79 / 1e6, - max_tokens=8192, - has_function_call_api=True, - ), ChatModelInfo( name=GroqModelName.MIXTRAL_8X7B, provider_name=ModelProviderName.GROQ,