Source: https://seerepo.com/r/Dao-AILab/flash-attention/built
Markdown source: https://seerepo.com/r/Dao-AILab/flash-attention/built.md
Description: flash-attention is a Python repo. It uses AI through apache-tvm-ffi, NVIDIA, quack-kernels, torch-c-dlpack-ext, triton-windows, and 3 more.

# What is flash-attention built with?

flash-attention is a Python repo. It uses AI through apache-tvm-ffi, NVIDIA, quack-kernels, torch-c-dlpack-ext, triton-windows, and 3 more. It has 139,521 written lines, with 80 test files.

Read on November 16, 2026, at commit 328dd7b.

Repository: https://github.com/Dao-AILab/flash-attention

## Backend

### AI

- **[apache-tvm-ffi](https://seerepo.com/uses/apache-tvm-ffi)**
- **[NVIDIA](https://seerepo.com/uses/nvidia)**
- **quack-kernels**
- **torch-c-dlpack-ext**
- **triton-windows**
- **[torch](https://seerepo.com/uses/torch)**: 86 files
- **[einops](https://seerepo.com/uses/einops)**: 25 files
- **[triton](https://seerepo.com/uses/triton)**: 10 files

## Support

### Tools

- **ninja**
- **packaging**
- **typing-extensions**: 1 file

### Other

- **win32**

## Other repos with a similar stack

- [NVIDIA/TensorRT-LLM](https://seerepo.com/r/NVIDIA/TensorRT-LLM/built): apache-tvm-ffi, NVIDIA, quack-kernels, torch-c-dlpack-ext, torch, einops, triton · [Compare](https://seerepo.com/compare/Dao-AILab/flash-attention/vs/NVIDIA/TensorRT-LLM)
- [vllm-project/vllm](https://seerepo.com/r/vllm-project/vllm/built): apache-tvm-ffi, NVIDIA, quack-kernels, torch, einops, triton · [Compare](https://seerepo.com/compare/Dao-AILab/flash-attention/vs/vllm-project/vllm)
- [sgl-project/sglang](https://seerepo.com/r/sgl-project/sglang/built): apache-tvm-ffi, NVIDIA, quack-kernels, torch-c-dlpack-ext, torch, einops · [Compare](https://seerepo.com/compare/Dao-AILab/flash-attention/vs/sgl-project/sglang)
- [state-spaces/mamba](https://seerepo.com/r/state-spaces/mamba/built): apache-tvm-ffi, quack-kernels, torch, einops, triton · [Compare](https://seerepo.com/compare/Dao-AILab/flash-attention/vs/state-spaces/mamba)
- [tile-ai/tilelang](https://seerepo.com/r/tile-ai/tilelang/built): apache-tvm-ffi, NVIDIA, torch-c-dlpack-ext, torch, einops · [Compare](https://seerepo.com/compare/Dao-AILab/flash-attention/vs/tile-ai/tilelang)
- [onyx-dot-app/onyx](https://seerepo.com/r/onyx-dot-app/onyx/built): NVIDIA, torch, einops, triton · [Compare](https://seerepo.com/compare/Dao-AILab/flash-attention/vs/onyx-dot-app/onyx)

## Frequently asked questions

### What language is flash-attention written in?

flash-attention is written in Python, with 139,521 written lines.

### Does flash-attention use AI?

Yes. flash-attention uses AI through apache-tvm-ffi, NVIDIA, quack-kernels, torch-c-dlpack-ext, triton-windows, torch, einops, and triton.

### Does flash-attention have tests?

Yes. flash-attention has 80 test files.

### Which repos have a stack like flash-attention's?

The ones that share most of its stack: NVIDIA/TensorRT-LLM, vllm-project/vllm, and sgl-project/sglang.

## What is your repo built with?

Any public repo, with no account: https://seerepo.com/r/owner/repo/built, with the owner and name it has on GitHub.
