vllm-project/vllm
Maintainers
vLLM is governed by vllm-project/vllm/.github/CODEOWNERS plus the recurring committer cohort visible in git log. This page summarizes ownership by subsystem so contributors know who to tag for review.
Core engine and configuration
| Area | Owners |
|---|---|
vllm/config/ (the VllmConfig family) |
@WoosukKwon @youkaichao @robertgshaw2-redhat @mgoin @tlrmchlsmth @houseroad @hmellor @yewentao256 @ProExpertProg |
vllm/config/cache.py |
@heheda12345 |
vllm/v1/core/ (scheduler + KV cache manager) |
@WoosukKwon @robertgshaw2-redhat @njhill @ywang96 @alexm-redhat @heheda12345 @ApostaC @orozery |
vllm/v1/engine/ |
@njhill |
vllm/v1/executor/ |
@njhill |
vllm/v1/worker/ |
@njhill |
vllm/v1/worker/gpu/ (model runner V2) |
@WoosukKwon @njhill |
vllm/v1/sample/ |
@22quinn @houseroad @njhill |
vllm/v1/kv_cache_interface.py |
@heheda12345 |
vllm/v1/kv_offload/ |
@ApostaC @orozery |
Attention
| Area | Owners |
|---|---|
vllm/v1/attention/ and vllm/model_executor/layers/attention/ |
@LucasWilkinson @MatthewBonanni |
vllm/v1/attention/backends/mla/ |
@pavanimajety |
vllm/v1/attention/backends/flashinfer.py |
@mgoin @pavanimajety @vadiklyutiy |
vllm/v1/attention/backends/triton_attn.py |
@tdoublep |
vllm/v1/attention/backends/gdn_attn.py |
@ZJY0516 @vadiklyutiy |
vllm/vllm_flash_attn |
@LucasWilkinson @MatthewBonanni |
| ROCm attention backends | @gshtras @tjtanaa |
Quantization, kernels, compilation
| Area | Owners |
|---|---|
vllm/model_executor/layers/quantization/ |
@mgoin @robertgshaw2-redhat @tlrmchlsmth @yewentao256 @pavanimajety |
vllm/model_executor/layers/fused_moe/ |
@mgoin @pavanimajety |
vllm/model_executor/layers/mamba/ |
@tdoublep @tomeras91 |
vllm/model_executor/layers/mamba/gdn_linear_attn.py |
@tdoublep @ZJY0516 @vadiklyutiy |
vllm/model_executor/layers/rotary_embedding.py |
@vadiklyutiy |
vllm/model_executor/layers/fla/ |
@ZJY0516 @vadiklyutiy |
vllm/model_executor/layers/batch_invariant.py |
@yewentao256 |
vllm/model_executor/layers/pooler/ |
@noooop |
vllm/model_executor/model_loader/ |
@22quinn |
vllm/compilation/ |
@zou3519 @youkaichao @ProExpertProg @BoyuanFeng @vadiklyutiy |
vllm/ir/ |
@ProExpertProg |
vllm/kernels/ |
@ProExpertProg @tjtanaa |
vllm/kernels/helion/ |
@ProExpertProg @zou3519 |
CMakeLists.txt |
@tlrmchlsmth @LucasWilkinson |
Multimodal
| Area | Owners |
|---|---|
vllm/multimodal/ |
@DarkLight1337 @ywang96 @NickLucche @tjtanaa |
LoRA
| Area | Owners |
|---|---|
vllm/lora/, tests/lora/ |
@jeejeelee |
Distributed / KV transfer
| Area | Owners |
|---|---|
vllm/distributed/kv_transfer/ |
@NickLucche @ApostaC @orozery @xuechendi |
vllm/v1/worker/kv_connector_model_runner_mixin.py |
@orozery @NickLucche |
vllm/v1/worker/gpu/kv_connector.py |
@orozery |
Speculative decoding and structured output
| Area | Owners |
|---|---|
vllm/v1/spec_decode/ |
@benchislett @luccafong @MatthewBonanni |
vllm/v1/structured_output/ |
@mgoin @russellb @aarnphm @benchislett |
Frontends
| Area | Owners |
|---|---|
vllm/entrypoints/openai/ |
@aarnphm @chaunceyjiang @DarkLight1337 @russellb |
vllm/entrypoints/openai/realtime/ |
@njhill |
vllm/entrypoints/openai/speech_to_text/ |
@NickLucche |
vllm/entrypoints/anthropic/ |
@mgoin @DarkLight1337 |
vllm/entrypoints/cli/ |
@hmellor @mgoin @DarkLight1337 @russellb |
vllm/entrypoints/mcp/ |
@heheda12345 |
vllm/entrypoints/sagemaker/ |
@DarkLight1337 |
vllm/entrypoints/serve/ |
@njhill |
vllm/entrypoints/pooling/ |
@noooop |
vllm/entrypoints/llm.py |
@DarkLight1337 |
vllm/entrypoints/chat_utils.py |
@DarkLight1337 |
vllm/sampling_params.py |
@njhill @NickLucche |
vllm/pooling_params.py |
@noooop @DarkLight1337 |
vllm/tokenizers/ |
@DarkLight1337 @njhill |
vllm/renderers/ |
@DarkLight1337 @njhill |
Tool calling and reasoning
| Area | Owners |
|---|---|
vllm/reasoning/ |
@aarnphm @chaunceyjiang @sfeng33 @bbrowning |
vllm/tool_parsers/ |
@aarnphm @chaunceyjiang @sfeng33 @bbrowning |
vllm/parser/ |
@aarnphm @chaunceyjiang @sfeng33 @bbrowning |
Platforms
| Platform | Owners |
|---|---|
| CPU | @bigPYJ1151 @xuechendi (vllm/v1/worker/cpu*, csrc/cpu, vllm/platforms/cpu.py, cmake/cpu_extension.cmake, docker/Dockerfile.cpu) |
| Intel XPU | @jikunshang @xuechendi (vllm/v1/worker/xpu*, vllm/platforms/xpu.py, docker/Dockerfile.xpu) |
| ROCm | @gshtras @tjtanaa (rocm-tagged files across the tree) |
| TPU | @NickLucche (vllm/v1/worker/tpu*, vllm/platforms/tpu.py, vllm/v1/sample/tpu/) |
Observability
| Area | Owners |
|---|---|
vllm/v1/metrics/, vllm/tracing.py, vllm/config/observability.py, vllm/config/kv_events.py, vllm/distributed/kv_events.py |
@markmc |
Documentation, lint, infra
| Area | Owners |
|---|---|
docs/, mkdocs.yaml, .readthedocs.yaml, requirements/docs.txt |
@hmellor |
.markdownlint.yaml, .pre-commit-config.yaml, tools/pre_commit/ |
@hmellor |
Model-family ownership
| Family | Owners |
|---|---|
Mistral / Mixtral / Pixtral / Voxtral and vllm/tokenizers/mistral.py |
@patrickvonplaten |
| Nemotron | @tomeras91 |
| Qwen | @sighingnow @vadiklyutiy |
Transformers backend (vllm/model_executor/models/transformers/) |
@hmellor |
deepseek_mtp.py |
@luccafong |
Security
| Area | Owners |
|---|---|
SECURITY.md, docs/usage/security.md, docs/contributing/vulnerability_management.md |
@russellb |
Highly active recurring contributors (90 days, by commits)
From git shortlog -sn --since=2026-01-30:
- Cyrus Leung (DarkLight1337)
- Harry Mellor
- Roger Wang
- Lu Fang
- Cody Yu
- Lily Liu
- Wentao Ye
- Andreas Karatzas
- Tyler Michael Smith
- Lucas Wilkinson
- Robert Shaw
- Aaron Pham
- Russell Bryant
- Michael Goin
- Nick Hill
The full picture is at git shortlog -sn (2,400+ unique committers). For up-to-date governance, the vLLM governance docs and the project Slack are authoritative.
Built by Factory AutoWiki from public repository content. It is a generated preview for codebase exploration, not source-maintained documentation.