- 16 Aug, 2025 16 commits
-
-
User will correctly get estimated prompt usage when upstream returns either zero or nothing.
funnycups committed -
Feats: use the types of gjson,the error expection,the invert and the key missing process of the condition
Calcium-Ion committed -
CaIon committed
-
Nekohy committed
-
Nekohy committed
-
Nekohy committed
-
CaIon committed
-
CaIon committed
-
CaIon committed
-
CaIon committed
-
feats: the flexable params override and compatible format
Calcium-Ion committed -
CaIon committed
-
feat: use proxy HTTP client fetch models
Calcium-Ion committed -
feat(relay): add OpenAI gpt-4.1 o3 o4 gpt-image-1 models
Calcium-Ion committed -
Nekohy committed
-
Nekohy committed
-
- 15 Aug, 2025 24 commits
-
-
CaIon committed
-
CaIon committed
-
CaIon committed
-
CaIon committed
-
CaIon committed
-
CaIon committed
-
wzxjohn committed
-
wzxjohn committed
-
CaIon committed
-
CaIon committed
-
Fix reasoning adaptor for openrouter
Calcium-Ion committed -
CaIon committed
-
CaIon committed
-
CaIon committed
-
CaIon committed
-
fix: prevent OpenRouter cache calculation with custom model ratios
Calcium-Ion committed -
yyhhyyyyyy committed
-
CaIon committed
-
手机客户端下拉菜单采用导航栏的样式
同語 committed -
refactor: Introduce pre-consume quota and unify relay handlers
Calcium-Ion committed -
CaIon committed
-
CaIon committed
-
CaIon committed
-
# Conflicts: # dto/openai_image.go
CaIon committed
-