trymirai/uzu
A Rust inference engine enabling zero-latency, privacy-preserving on-device AI model execution.

Velocity · 7d
+4.6
★ / day
Trend
→steady
star history
Uzu is a high-performance inference engine written in Rust for running AI models directly within applications. It provides bindings for Python, TypeScript, and Swift to deploy AI locally with full data privacy and no inference costs. The engine leverages Metal for GPU acceleration on Apple platforms and targets use cases including LLMs and text-to-speech.