Candle

Minimalist ML framework for Rust with GPU support

D
Score: 48/100
Type
Execution
jit
Interface
sdk

About

Candle is a minimalist ML framework for Rust focused on performance and ease of use. Created by Hugging Face, it supports CUDA and Metal backends with a focus on inference. Great for embedding LLMs in Rust applications.

Performance

300ms
Cold Start
200MB
Base Memory
50ms
Startup Overhead

Last Verified

Date: Jan 18, 2026
Method: manual test

Manually verified

Languages

Rust

Details

Isolation
process
Maturity
stable
License
Apache-2.0

Links