# tracel-ai/burn

Burn is a next generation tensor library and Deep Learning Framework that doesn't compromise on flexibility, efficiency and portability.

Repository: https://github.com/tracel-ai/burn
Canonical: https://ross.abutalabs.com/products/burn
Homepage: https://burn.dev
Language: Rust
License: Apache-2.0
License Family: permissive
Topics: autodiff, deep-learning, machine-learning, rust, scientific-computing, ndarray, tensor, neural-network, pytorch, cross-platform, kernel-fusion, onnx, wasm, webgpu, cuda, metal, rocm, vulkan
Last push: 2026-08-26T20:21:09+00:00

## Health v2 (maintenance only)
Score: 89/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 71, longevity 100
- inputs: {"age_days": 1507, "days_push": 7, "days_rel": 118, "gap_med": 57.0, "n_releases_24m": 11}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 15816, forks 1026 (observed 2026-08-28T04:11:14.135660+00:00)

## What it is
Burn is a Rust-based tensor library and deep learning framework supporting training and inference through a unified API. It JIT-compiles tensor operations with automatic kernel fusion, running on CUDA, Metal, WebGPU, Vulkan, ROCm, and CPU across servers, desktops, and mobile devices.

## Use cases
- train neural networks in rust
- run deep learning inference on any device
- deploy models without python export step
- on-device model personalization and federated learning
- run models in the browser with webgpu and wasm
- port pytorch models via onnx to rust
- gpu kernel fusion for faster inference

## When to choose
- you want a single rust codebase for training and production inference
- you need portable deployment across gpu vendors, mobile, and wasm
- you want pytorch-like ergonomics with compiled performance

## When to avoid
- your team and tooling are python-centric and you don't need rust
- you rely on python-only ecosystem features like certain training libraries
- you need the broadest model zoo support available today

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-training, llm-inference, gpu-computing, wasm
- domain: deep-learning, machine-learning, gpu-computing, cross-platform
- platform: rust, wasm, cross-platform
- tags: tensor-library, autodiff, kernel-fusion, onnx, webgpu, pytorch-alternative, scientific-computing, cuda, gpu

## Member repositories
- tracel-ai/burn (main) score 89

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:14.135660+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:05:31.938549+00:00, confidence not recorded.
  - readme: https://github.com/tracel-ai/burn (fetched 2026-08-28T04:11:14.135660+00:00, sha 5de7ad292aaa)
  - homepage: https://burn.dev (fetched 2026-08-29T08:03:20.528367+00:00, sha 2991f31e852e)
  - registry_crates: https://crates.io/api/v1/crates/burn (fetched 2026-08-29T08:03:20.530929+00:00, sha fb3898cb6b3d)
- Data as of 2026-08-30T08:39:29.467469+00:00.