master
Autoparser - complete refactoring of parser architecture (#18675)
common/parser: add proper reasoning tag prefill reading (#20424)
chat : Seed OSS thinking + tool call support (#15552)
Tool call support (generic + native for Llama, Functionary, Hermes, Mistral, Firefunction, DeepSeek) w/ lazy grammars (#9639)
common : Generalized XML-style tool-call parsing with streaming support (GLM 4.5/4.6 + MiniMax M2 + SeedOSS + Kimi-K2 + Qwen3-Coder + Apriel-1.5 + Xiaomi-MiMo) (#16932)
common/parser: add GigaChatV3/3.1 models support (#19931)
common/autoparser : detect reasoning markers when enable_thinking changes system prompt (#20859)
Fix Kimi-K2 tool-call parsing issues (#17376)
PEG parser for LFM2 (#20251)
fix: tool call parsing for LFM2 and LFM2.5 models (#21242)
jinja : Add Mistral-Small-3.2-24B-Instruct-2506.jinja (#14349)
common : implement new jinja template engine (#18462)
chat : nemotron thinking & toolcalling support (#15676)
`server`: add `--reasoning-budget 0` to disable thinking (incl. qwen3 w/ enable_thinking:false) (#13771)
common/parser: fix handling of tool definition with missing properties key (#21128)
chat : Deepseek V3.1 reasoning and tool calling support (OpenAI Style) (#15533)
autoparser: support case of JSON_NATIVE with per-call markers (test case: Reka-Edge) (#21892)
chat: dedicated DeepSeek v3.2 parser + "official" template (#21785)
common : better align to the updated official gemma4 template (#21704)
chat : support Granite model reasoning and tool call (#14864)
chat : add Granite 4.0 chat template with correct tool_call role mapping (#20804)
llama : add jinja template for rwkv-world (#14665)
common : add parser for ministral/mistral large 3/devstral 2 (#17713)
gpt-oss: implement harmony parsing (#15181)
common : fix Step-3.5-Flash format detection and thinking support (#19635)
mtmd : add support for Voxtral (#14862)
chat : add parsing for solar-open-100b (#18540)