Source: https://seerepo.com/compare/Dao-AILab/flash-attention/vs/state-spaces/mamba
Markdown source: https://seerepo.com/compare/Dao-AILab/flash-attention/vs/state-spaces/mamba.md
Description: flash-attention and mamba share 5 technologies: apache-tvm-ffi, quack-kernels, torch, einops, and triton. Both are written in Python.

# flash-attention vs mamba

flash-attention and mamba share 5 technologies: apache-tvm-ffi, quack-kernels, torch, einops, and triton. Both are written in Python.

- [Dao-AILab/flash-attention](https://seerepo.com/r/Dao-AILab/flash-attention): Fast and memory-efficient exact attention
- [state-spaces/mamba](https://seerepo.com/r/state-spaces/mamba): Mamba SSM architecture

## What they share and what sets them apart

- In common: apache-tvm-ffi, quack-kernels, torch, einops, triton
- Only in flash-attention: NVIDIA, torch-c-dlpack-ext, triton-windows
- Only in mamba: transformers

## Side by side

| | flash-attention | mamba |
|---|---|---|
| Size | 139,521 lines written, 150 code files | 28,235 lines written, 43 code files |
| Who writes it | 1.7% of its commits signed by AI. With Claude Code, Cursor, and GitHub Copilot | 0.5% of its commits signed by AI. With GitHub Copilot |
| Activity | Last push November 16, 2026, 1,613 commits, 25,304 stars | Last push July 22, 2026, 219 commits, 18,976 stars |
| Tests | 80 test files. 32% of its code files have a test that imports them | 11 test files. 51.2% of its code files have a test that imports them |

- Size: Lines written by people and agents, without generated ones; code files, without tests.
- Who writes it: A commit counts as AI's when it carries its signature: a Co-Authored-By line, its author name or the merge of its branch.
- Activity: From GitHub and its history, on the day SeeRepo read it.
- Tests: The files that test the code, and the share of code files a test imports.

## Frequently asked questions

### Which is bigger, flash-attention or mamba?

flash-attention, with 139,521 lines written; mamba has 28,235.

The proof: [Dao-AILab/flash-attention's code at commit 328dd7b](https://github.com/Dao-AILab/flash-attention/tree/328dd7b5a6905c75500a9bce334ac012763409a9)

### Which has more GitHub stars, flash-attention or mamba?

flash-attention, with 25,304 stars; mamba has 18,976.

The proof: [Dao-AILab/flash-attention on GitHub](https://github.com/Dao-AILab/flash-attention)

### Which has more tests, flash-attention or mamba?

flash-attention, with 80 test files; mamba has 11.

The proof: [Dao-AILab/flash-attention's code at commit 328dd7b](https://github.com/Dao-AILab/flash-attention/tree/328dd7b5a6905c75500a9bce334ac012763409a9)

### Which was updated more recently, flash-attention or mamba?

flash-attention: its last push was on November 16, 2026, and mamba's on July 22, 2026.

The proof: [Dao-AILab/flash-attention's commits on GitHub](https://github.com/Dao-AILab/flash-attention/commits)

### Which has more commits signed by AI, flash-attention or mamba?

flash-attention: an AI signed 1.7% of its commits, with Claude Code, Cursor, and GitHub Copilot; in mamba, 0.5%.

The proof: [the commits each tool signed in Dao-AILab/flash-attention](https://seerepo.com/r/Dao-AILab/flash-attention/ai)

## Sources

- Dao-AILab/flash-attention: Read on November 16, 2026, at commit 328dd7b (https://github.com/Dao-AILab/flash-attention/tree/328dd7b5a6905c75500a9bce334ac012763409a9)
- state-spaces/mamba: Read on November 6, 2026, at commit e9594ce (https://github.com/state-spaces/mamba/tree/e9594ce1c732d97440f0332fdc43170a2294dbfa)

## 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.
