Preview of magnitudedev/magnitude

magnitudedev/magnitude

Open source inference engine that tunes its kernels on your hardware to find the fastest way to run each model. 2x faster than Ollama. Works on Apple Silicon, NVIDIA, AMD, Intel, or just a CPU.

★ 7.6K Stars
⑂ 541 Forks
○ 7.6K Watchers
✖ 48 Issues
Rust

Last push:

Created:

Loading additional information…