{"slug":"vllm","name":"vllm","category":"local-inference","categoryName":"Local inference & serving","audience":["ai-era","for-ai"],"primary":"ai-era","openClosed":"open","pricing":"open-source","license":"Apache-2.0","maintained":"yes","officialUrl":"https://vllm.ai","repo":"https://github.com/vllm-project/vllm","latestVersion":"v0.31.0","latestRelease":"2026-10-05","stars":94910,"starsAsOf":"2026-10-22","description":"A high-throughput and memory-efficient inference and serving engine for LLMs","descriptionZh":"高吞吐 LLM 推理与服务引擎","callMethods":["API","CLI"],"platforms":[],"addedAt":"2026-10-10","updatedAt":"2026-10-22","topics":["amd","blackwell","cuda","deepseek","deepseek-v3","gpt","gpt-oss","inference"],"language":"Python","lastPush":"2026-10-22","formerly":null,"packages":[{"registry":"pypi","name":"vllm","url":"https://pypi.org/project/vllm/","versionsUrl":"https://pypi.org/project/vllm/#history","latestVersion":"0.31.0","latestDate":"2026-10-05"}],"recentVersions":[{"version":"v0.31.0","date":"2026-10-05","prerelease":false,"url":"https://github.com/vllm-project/vllm/releases/tag/v0.31.0","source":"GitHub Release"},{"version":"v0.30.0","date":"2026-09-22","prerelease":false,"url":"https://github.com/vllm-project/vllm/releases/tag/v0.30.0","source":"GitHub Release"},{"version":"v0.29.0","date":"2026-09-09","prerelease":false,"url":"https://github.com/vllm-project/vllm/releases/tag/v0.29.0","source":"GitHub Release"},{"version":"v0.28.0","date":"2026-08-26","prerelease":false,"url":"https://github.com/vllm-project/vllm/releases/tag/v0.28.0","source":"GitHub Release"},{"version":"v0.27.1","date":"2026-08-11","prerelease":false,"url":"https://github.com/vllm-project/vllm/releases/tag/v0.27.1","source":"GitHub Release"}],"alternatives":["sglang","gpustack","ollama","lmcache","tensorrt-llm","cherry-studio","llm-d","mediapipe"],"stacks":["fine-tune-open-models","airgapped-intranet"],"scenarios":["switch-models","run-models-locally","use-ai-compliantly-in-china","learn-ai","enterprise-private-ai"],"problems":["vllm-cuda-out-of-memory","use-llms-compliantly-in-china","deploy-dify-self-hosted"],"reasonZh":"企业私有化推理事实标准；2,241 名提交者","sources":["profile","priority","oss","github-pool"],"record":"https://aimap.page/api/v1/software/vllm.json","pages":{"en":"https://aimap.page/en/software/vllm/","zh-hans":"https://aimap.page/zh-hans/software/vllm/"}}