MLC-LLM
Compile-and-serve LLMs on Metal, Vulkan, WebGPU, and phones — no CUDA required.
Fit.
Catalog snapshot
- Platforms
- Linux, Macos, Windows, Ios, Android, Web
- Operating systems
- Linux, macOS, Windows, iOS, Android, Web
- Ram floor
- 8 GB
- Gpu
- Optional · 6 GB VRAM helps
- Docker
- No
- Setup
- Advanced
- Deploy as
- Local, Self-hosted
- Commercial use
- Allowed
Runs chat models on phones and in browsers — the deployment surface is the point.
Project.
Fetched 1
About.
MLC-LLM compiles models to run anywhere ML compilation reaches: Apple Silicon, AMD, Android, iOS, and even WebGPU in the browser, with an OpenAI-compatible REST server. Apache-2.0. The escape hatch from CUDA-only serving.
ML compilation · Metal, Vulkan, and WebGPU backends · REST server · iOS and Android targets
