Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
Commits on main
create: select the qwen3.5 parser and renderer for Qwen3.5/Next
d47859c Jesse Gross committed 5 months ago
llm: allow iGPU mmproj offload with fit padding (#16996)
a6293eb Daniel Hiltgen committed 4 months ago
server: apply format constraint for all thinking parsers when think=false (#15901)
892e7f6 Arkadeep Dutta committed 4 months ago
launch: disable claude code telemetry by default (#17061)
87b6421 Parth Sareen committed 4 months ago
llama: clean up dead code from llama-server work (#17007)
7b22ac9 Daniel Hiltgen committed 4 months ago
discover: fall back to standard CUDA when the JetPack runner is absent (#16949)
624cada Kevin Park committed 4 months ago
launch: check for min version for hermes desktop (#16912)
ada1eb5 Bruce MacDonald committed 4 months ago
tools: ignore braces inside JSON strings when detecting tool call end (#16937)
32a97b7 Aditya Aggarwal committed 4 months ago
MLX: wire up scheduler selected context size for ps (#16918)
d26a585 Daniel Hiltgen committed 4 months ago
parser/renderer: add Ornith 9B renderer/parser support (#16920)
2e474c9 Parth Sareen committed 4 months ago
launch: update hermes install urls to official (#16913)
2cb2c53 Bruce MacDonald committed 4 months ago