GitHub · Rust
basecompute/superfluid
Superfluid is an open-source server that runs LLMs on your own hardware. It serves a model from one machine, or from several machines that act as one server. It supports llama.cpp, MLX, and BaseRT, and exposes OpenAI-, Anthropic-, and Ollama-compatible API endpoints.
- Stars
- 31
- Forks
- 4
- Stars gained in 24 hours
- 0
- Forks gained in 24 hours
- 0
Topics: anthropic-api, apple-silicon, claude-code, coding-agent, continuous-batching, inference-server, llama-cpp, llm-inference, llm-serving, local-llm
Latest data point: Nov 20, 2026 · 31 stars · 4 forks