{"slug":"flash-attention","name":"flash-attention","category":"local-inference","categoryName":"Local inference & serving","audience":["ai-era"],"primary":"ai-era","openClosed":"open","pricing":"open-source","license":"BSD-3-Clause","maintained":"yes","officialUrl":"https://github.com/Dao-AILab/flash-attention","repo":"https://github.com/Dao-AILab/flash-attention","latestVersion":"v2.8.3.post2","latestRelease":"2026-10-14","stars":25193,"starsAsOf":"2026-10-23","description":"Fast and memory-efficient exact attention","descriptionZh":"快速显存高效注意力内核","callMethods":["SDK"],"platforms":[],"addedAt":"2026-10-10","updatedAt":"2026-10-23","topics":[],"language":"Python","lastPush":"2026-10-23","formerly":null,"packages":[],"recentVersions":[{"version":"fa4-v4.0.0.beta34","date":"2026-10-07","prerelease":true,"url":"https://github.com/Dao-AILab/flash-attention/releases/tag/fa4-v4.0.0.beta34","source":"GitHub Release"},{"version":"fa4-v4.0.0.beta33","date":"2026-09-30","prerelease":true,"url":"https://github.com/Dao-AILab/flash-attention/releases/tag/fa4-v4.0.0.beta33","source":"GitHub Release"},{"version":"fa4-v4.0.0.beta32","date":"2026-09-23","prerelease":true,"url":"https://github.com/Dao-AILab/flash-attention/releases/tag/fa4-v4.0.0.beta32","source":"GitHub Release"},{"version":"fa4-v4.0.0.beta31","date":"2026-09-16","prerelease":true,"url":"https://github.com/Dao-AILab/flash-attention/releases/tag/fa4-v4.0.0.beta31","source":"GitHub Release"},{"version":"fa4-v4.0.0.beta30","date":"2026-09-09","prerelease":true,"url":"https://github.com/Dao-AILab/flash-attention/releases/tag/fa4-v4.0.0.beta30","source":"GitHub Release"}],"alternatives":["ollama","llama-cpp","vllm","cherry-studio","localai","exo","sglang","mlx"],"stacks":[],"scenarios":["switch-models","run-models-locally","use-ai-compliantly-in-china","learn-ai"],"problems":[],"reasonZh":"开源大库收录：快速显存高效注意力内核","sources":["oss","github-oss","github-pool"],"record":"https://aimap.page/api/v1/software/flash-attention.json","pages":{"en":"https://aimap.page/en/software/flash-attention/","zh-hans":"https://aimap.page/zh-hans/software/flash-attention/"}}