High-performance LLM inference on Apple Silicon using MLX and vLLM
vLLM Metal is a plugin that enables vLLM to run on Apple Silicon Macs using MLX as the primary compute backend. It unifies MLX and PyTorch under a single lowering path.
Documentation: https://docs.vllm.ai/projects/vllm-metal/en/latest/
Latest News 🔥
- [2026/08] vLLM Metal now uses M5 NAX tensor units to accelerate MHA, GQA, and MQA prefill.
- [2026/08] Qwen3.8 now runs on Metal!
mlx-community/Qwen3.8-27B-8bitserves a 27B hybrid SDPA + GDN linear model on a single Apple Silicon Mac. - [2026/04] We released the new version v0.2.0! Unified paged varlen Metal kernel is now the default attention backend. 83x TTFT, 3.6x throughput compared to v0.1.0.
Upstream vLLM supplies the API server, scheduler, and paged block manager; mlx_lm supplies the token-wise model layers; vllm-metal owns the request-aware attention path — the paged varlen kernel, M5 NAX prefill, and speculative decoding.
- macOS 15 (Sequoia) or later, on Apple Silicon
- Native arm64 Python 3.12. Rosetta/x86_64 Python is not supported.
vllm-metal supports a growing set of models on Apple Silicon. See the full matrix in docs/supported_models.md.
curl -fsSL https://raw.githubusercontent.com/vllm-project/vllm-metal/main/install.sh | bashUsing the install script above, the following will be installed under the ~/.venv-vllm-metal directory (the default).
- vllm-metal plugin
- vllm core
- Related libraries
If you run source ~/.venv-vllm-metal/bin/activate, the vllm CLI becomes available and you can access the vLLM right away.
For how to use the vllm CLI, please refer to the official vLLM guide.
https://docs.vllm.ai/en/latest/cli/