Gitinbox/forge
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: register DeepSeek deepseek-chat/deepseek-reasoner in model catalog
CI / API - lint + tests (push) Has been cancelled
Details
CI / Web - typecheck + build (push) Has been cancelled
Details
CodeQL / Analyze (javascript-typescript) (push) Has been cancelled
Details
CodeQL / Analyze (python) (push) Has been cancelled
Details

Browse Source 
- Add openai:deepseek-chat (V3) and openai:deepseek-reasoner (R1) to CHAT_MODELS
- DeepSeek is OpenAI-compatible; endpoint routed via OPENAI_BASE_URL env
- Prices per DeepSeek official: chat $0.28/$0.42, reasoner $0.55/$2.19 per 1M tokens
This commit is contained in:
figmar
2026-08-10 20:21:24 +08:00
parent f311f29b42
commit 78c369e651
1 changed files with 3 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
apps/api/forge/model_catalog.py
+3
View File
@@ -46,6 +46,9 @@ CHAT_MODELS: list[ModelInfo] = [
# Google
ModelInfo("google_genai:gemini-1.5-flash", "gemini-1.5-flash", "Google", "1M", True, True, 0.075, 0.3),
ModelInfo("google_genai:gemini-2.5-flash", "gemini-2.5-flash", "Google", "1M", True, True, 0.3, 2.5),
# DeepSeek (OpenAI-compatible endpoint; base URL set via OPENAI_BASE_URL env)
ModelInfo("openai:deepseek-chat", "deepseek-chat", "DeepSeek", "128k", True, False, 0.28, 0.42),
ModelInfo("openai:deepseek-reasoner", "deepseek-reasoner", "DeepSeek", "128k", True, False, 0.55, 2.19),
# Offline / test - runs with no provider credentials; never priced (see catalog_prices).
ModelInfo("fake:echo", "fake (offline test)", "Local", "-", True, False, 0.0, 0.0),
]