Muzic: Music Understanding and Generation with Artificial Intelligence

Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

Split text into semantic chunks, up to a desired chunk size. Supports calculating length by characters and tokens, and is callable from Rust and Python.

LLM Orchestrator built in Rust

Low rank adaptation (LoRA) for Candle.

Efficent platform for inference and serving local LLMs including an OpenAI compatible API server.

Instant, controllable, local pre-trained AI models in Rust

A blazing fast inference solution for text embeddings models