EnzymeAD/Enzyme
High-performance automatic differentiation of LLVM and MLIR.
Operates as an LLVM/MLIR compiler plugin that synthesizes gradient functions at the IR level, enabling differentiation of optimized code and foreign libraries without source modification. Supports reverse-mode AD across GPUs (CUDA/ROCm), parallel paradigms (OpenMP, MPI, Julia), and provides bindings for Julia and Rust ecosystems through simple `__enzyme_autodiff` API calls.
1,558 stars. Actively maintained with 37 commits in the last 30 days.
Stars
1,558
Forks
154
Language
LLVM
License
—
Category
Last pushed
Mar 13, 2026
Commits (30d)
37
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/EnzymeAD/Enzyme"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
Oxen-AI/Oxen
Lightning fast data version control system for structured and unstructured machine learning...
LaurentMazare/tch-rs
Rust bindings for the C++ api of PyTorch.
SunDoge/dlpark
A Rust Library for High-Performance Tensor Exchange with Python
strake-data/strake
The Data Layer for AI. A high-performance federated SQL engine that gives AI agents governed,...
Photoroom/datago
A natively parallel dataloader for Python, written in Rust. Serving data at GB/s speeds, while...