Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
Commits on main
server/openai: retry compaction after context overflow (#18324)
1b45aa1 Parth Sareen committed 2 months ago
app: preserve Codex configuration across host changes (#18247)
9ef6c19 Parth Sareen committed 2 months ago
openai: use tsc_ prefix for tool search call IDs (#18296)
3b5ab1f Jeffrey Morgan committed 2 months ago
openai: preserve images through response compaction (#18245)
b043d89 Parth Sareen committed 2 months ago
app: route Codex auto review through selected model (#18246)
43e6670 Parth Sareen committed 2 months ago
ci: bound and cache the macOS MLX payload build (#18240)
d3efc63 Daniel Hiltgen committed 2 months ago
mlxrunner: compile structured output as xgrammar structural tags
b68365a Jesse Gross committed 2 months ago
mlxrunner: enable speculative decoding under structured output
4986e92 Jesse Gross committed 2 months ago
ci: rebuild MLX macOS test payloads the release can't supply (#18219)
59fe23d Daniel Hiltgen committed 2 months ago
mlxrunner: shared audio decoding and chunking for audio models
3ffc9a6 Jesse Gross committed 2 months ago
mlxrunner: capture MLX errors in a single buffer read after every call
c36adeb Jesse Gross committed 2 months ago
fix(docs): correct typos found during code review (#17579)
e37a00a Marcel Petrick committed 2 months ago