[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-9572":3},{"id":4,"name":5,"fullName":6,"owner":5,"repo":5,"description":7,"homepage":8,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":15,"forks30d":15,"starsTrendScore":19,"compositeScore":20,"rankGlobal":9,"rankLanguage":9,"license":21,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":22,"hasPages":24,"topics":25,"createdAt":9,"pushedAt":9,"updatedAt":32,"readmeContent":33,"aiSummary":34,"trendingCount":15,"starSnapshotCount":15,"syncStatus":35,"lastSyncTime":36,"discoverSource":37},9572,"modular","modular\u002Fmodular","The Modular Platform (includes MAX & Mojo)","https:\u002F\u002Fdocs.modular.com\u002F",null,"Mojo",26323,2839,263,761,0,1,45,212,18,102,"Other",false,"main",true,[26,27,28,29,5,30,31],"ai","language","machine-learning","max","mojo","programming-language","2026-06-12 04:00:45","\u003Cdiv align=\"center\">\n    \u003Cimg src=\"https:\u002F\u002Fmodular-assets.s3.amazonaws.com\u002Fimages\u002Fmodular_github_logo_bg.png\">\n\n[About Modular] | [Get started] | [API docs] | [Contributing]\n  | [Changelog] | [MAX Model Development]\n\u003C\u002Fdiv>\n\n---\n\n🤝 Join our [monthly community meetings][public-com-meet-doc]!\n\n# Modular Platform\n\n> A unified platform for AI development and deployment, including **MAX**🧑‍🚀 and\n**Mojo**🔥.\n\nThe Modular Platform is an open and fully-integrated suite of AI libraries\nand tools that accelerates model serving and scales GenAI deployments. It\nabstracts away hardware complexity so you can run the most popular open\nmodels with industry-leading GPU and CPU performance without any code changes.\n\n![](https:\u002F\u002Fdocs.modular.com\u002Fimages\u002Fmodular-container-stack.png?20250513)\n\n## Get started\n\nYou don't need to clone this repo.\n\nYou can install Modular as a `pip` or `conda` package and then start an\nOpenAI-compatible endpoint with a model of your choice.\n\nTo get started with the Modular Platform and serve a model using the MAX\nframework, see [the quickstart guide](https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Fget-started).\n\n> [!NOTE]\n> **Nightly vs. stable releases**\n> If you cloned the repo and want a stable release, run\n  `git checkout modular\u002FvX.X` to match the version.\n> The `main` branch tracks nightly builds, while the `stable` branch matches\n  the latest released version.\n\nAfter your model endpoint is up and running, you can start sending the model\ninference requests using\n[our OpenAI-compatible REST API](https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Frest-api\u002F).\n\nExplore all the models you can deploy with Modular in our\n[Model Library](https:\u002F\u002Fwww.modular.com\u002Fmodels).\n\n## Deploy our container\n\nThe MAX container is our Kubernetes-compatible Docker container for convenient\ndeployment, which uses the MAX framework's built-in inference server. We have\nseparate containers for NVIDIA and AMD GPU environments, and a unified container\nthat works with both.\n\nFor example, you can start a container for an NVIDIA GPU with this command:\n\n```sh\ndocker run --gpus=1 \\\n    -v ~\u002F.cache\u002Fhuggingface:\u002Froot\u002F.cache\u002Fhuggingface \\\n    -p 8000:8000 \\\n    modular\u002Fmax-nvidia-full:latest \\\n    --model google\u002Fgemma-3-27b-it\n```\n\nFor more information, see our [MAX container\ndocs](https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Fcontainer) or the [Modular Docker Hub\nrepository](https:\u002F\u002Fhub.docker.com\u002Fu\u002Fmodular).\n\n## About the repo\n\nWe're constantly open-sourcing more of the Modular Platform and you can find\nall of it in here. As of May, 2025, this repo includes over 450,000 lines of\ncode from over 6000 contributors, providing developers with production-grade\nreference implementations and tools to extend the Modular Platform with new\nalgorithms, operations, and hardware targets.\n\nIt's quite likely **the world's largest repository of open source CPU and GPU\nkernels**!\n\nHighlights include:\n\n- Mojo standard library: [\u002Fmojo\u002Fstdlib](mojo\u002Fstdlib)\n- MAX GPU and CPU kernels: [\u002Fmax\u002Fkernels](\u002Fmax\u002Fkernels) (Mojo kernels)\n- MAX inference server: [\u002Fmax\u002Fpython\u002Fmax\u002Fserve](\u002Fmax\u002Fpython\u002Fmax\u002Fserve)\n  (OpenAI-compatible endpoint)\n- MAX model pipelines: [\u002Fmax\u002Fpython\u002Fmax\u002Fpipelines](\u002Fmax\u002Fpython\u002Fmax\u002Fpipelines)\n  (Python-based graphs)\n- Code examples: [\u002Fmax\u002Fexamples](\u002Fmax\u002Fexamples) +\n  [\u002Fmojo\u002Fexamples](mojo\u002Fexamples)\n\nThis repo has two major branches:\n\n- The [`main`](https:\u002F\u002Fgithub.com\u002Fmodular\u002Fmodular\u002Ftree\u002Fmain) branch, which is\nin sync with the nightly build and subject to new bugs. Use this branch for\n[contributions](.\u002FCONTRIBUTING.md), or if you [installed the nightly\nbuild](https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Fpackages).\n\n- The [`stable`](https:\u002F\u002Fgithub.com\u002Fmodular\u002Fmodular\u002Ftree\u002Fstable) branch, which\nis in sync with the last stable released version of Mojo. Use the examples in\nhere if you [installed the stable\nbuild](https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Fpackages).\n\n## Contribute\n\nWe accept contributions to the [Mojo standard library](.\u002Fmojo), [MAX AI\nkernels](.\u002Fmax\u002Fkernels), [MAX model\narchitectures](\u002Fmax\u002Fpython\u002Fmax\u002Fpipelines\u002Farchitectures), code examples, Mojo\ndocs, and more.\n\nFirst, please read the [Contribution Guide](.\u002FCONTRIBUTING.md), and then refer\nto the following documentation about how to develop in the repo:\n\n- [`\u002Fmax\u002Fdocs`](\u002Fmax\u002Fdocs): Docs for developers working in the MAX framework\n  codebase.\n- [`\u002Fmojo\u002Fstdlib\u002Fdocs`](\u002Fmojo\u002Fstdlib\u002Fdocs): Docs for developers working in the\n  Mojo standard library.\n\nWe also welcome your bug reports. If you have a bug, please [file an issue\nhere](https:\u002F\u002Fgithub.com\u002Fmodular\u002Fmodular\u002Fissues\u002Fnew\u002Fchoose).\n\n## News & Announcements\n\n**[2026\u002F3]** [Modular Platform 26.2][26.2] delivers state-of-the-art image\ngeneration with over 4x speedup on FLUX.2 models, expanded hardware support for\nNVIDIA B300, Jetson Thor, DGX Spark, and AMD RDNA consumer GPUs, and Mojo\nlanguage upgrades that make it easier to write GPU kernels with AI coding\nagents.\n\n**[2026\u002F2]** We announced that [BentoML is joining Modular][bentoml-joins].\nWe are committed to building in the open and will be extending our support\nof open source AI with [Bento's own open project][bentoml-github].\nRead the answers in our [February 2026 AMA][bentoml-joins-ama] to learn more\nabout our plans.\n\n**[2026\u002F1]** [Modular Platform 26.1][26.1] graduates the MAX Python API out of\nexperimental with PyTorch-like eager mode and model.compile() for production,\nstabilizes the MAX LLM Book, and expands Apple silicon GPU support. Mojo gains\ncompile-time reflection, linear types, typed errors, and improved error messages\nas it progresses toward 1.0.\n\n**[2025\u002F12]** [The Path to Mojo 1.0][mojo-1.0] was officially announced\nwith a planned release in H1 2026 and tons of details on what to expect.\n\n**[2025\u002F12]** We hosted our [Inside the MAX Framework Meetup][dec-meetup]\nreintroducing the MAX framework and taking the community through upcoming\nchanges.\n\n**[2025\u002F11]** [Modular Platform 25.7][25.7] provides a fully open MAX Python\nAPI, expanded hardware support for NVIDIA Grace superchips, improved Mojo GPU\nprogramming experience, and much more.\n\n**[2025\u002F11]** We met with the community at\n[PyTorch 2025 + the LLVM Developers' Meeting][pytorch-llvm] to solicit\ncommunity input into how the Modular platform can reduce fragmentation and\nprovide a unified AI stack.\n\n**[2025\u002F09]** [Modular raises \\$250M][250-funding] to scale AI's unified compute\nlayer, bringing Modular's total raise to $380M at a $1.6B valuation.\n\n**[2025\u002F09]** [Modular Platform 25.6][25.6] delivers a unified compute layer\nspanning from laptops to datacenter GPUs, with industry-leading throughput on\nNVIDIA Blackwell (B200) and AMD MI355X.\n\n**[2025\u002F08]** [Modular Platform 25.5][25.5] introduces Large Scale Batch\nInference through a partnership with SF Compute + open source launch of the\nMAX Graph API and more.\n\n**[2025\u002F08]** We hosted our [Los Altos Meetup][la-meetup] featuring talks from\nChris Lattner on democratizing AI compute and Inworld AI on production voice AI.\n\n**[2025\u002F06]** [AMD partnership announced][amd] — Modular Platform now generally\navailable across AMD's MI300 and MI325 GPU portfolio.\n\n**[2025\u002F06]** [Modular Hack Weekend][hack-weekend] brought developers together\nto build custom kernels, model architectures, and PyTorch custom ops with\nMojo and MAX.\n\n**[2025\u002F05]** Over 100 engineers gathered at AGI House for our first\n[GPU Kernel Hackathon][hackathon], featuring talks from Modular and\nAnthropic engineers.\n\n---\n\n## Community & Events\n\nWe host regular meetups digitally and around the world. During these meetups\nwe share updates from the Modular team, feature community contributions, and\ninvite guest speakers to share their expertise, as well as answer community\nquestions.\n\nJoin us!\n\n| Channel               | Link                                            |\n|-----------------------|-------------------------------------------------|\n| 💬 Discord            | [discord.gg\u002Fmodular][discord]                   |\n| 💬 Forum              | [forum.modular.com][forum]                      |\n| 📅 Meetup Group       | [meetup.com\u002Fmodular-meetup-group][meetup-group] |\n| 🎥 Community Meetings | [Upcoming community calls][public-com-meet-doc] |\n\n**Upcoming events** will be posted on our [Meetup page][meetup-group] and\n[Discord][discord]. Community meeting recordings will be posted on our\n[YouTube][youtube].\n\n## Contact us\n\nIf you'd like to chat with the team and other community members, please send a\nmessage to our [Discord channel](https:\u002F\u002Fdiscord.gg\u002Fmodular) and [our\nforum board](https:\u002F\u002Fforum.modular.com\u002F).\n\n## License\n\nThis repository and its contributions are licensed under the Apache License v2.0\nwith LLVM Exceptions (see the LLVM [License](https:\u002F\u002Fllvm.org\u002FLICENSE.txt)).\nModular, MAX and Mojo usage and distribution are licensed under the\n[Modular Community License](https:\u002F\u002Fwww.modular.com\u002Flegal\u002Fcommunity).\n\n### Third party licenses\n\nYou are entirely responsible for checking and validating the licenses of third\nparties (i.e. Huggingface) for related software and libraries that are\ndownloaded.\n\n## Thanks to our contributors\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmodular\u002Fmodular\u002Fgraphs\u002Fcontributors\">\n  \u003Cimg src=\"https:\u002F\u002Fcontrib.rocks\u002Fimage?repo=modular\u002Fmodular\" \u002F>\n\u003C\u002Fa>\n\n\u003C!-- Link references -->\n\n\u003C!-- Header navigation links -->\n[About Modular]: https:\u002F\u002Fwww.modular.com\u002F\n[Get started]: https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Fget-started\n[API docs]: https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Fapi\n[Contributing]: .\u002FCONTRIBUTING.md\n[Changelog]: https:\u002F\u002Fdocs.modular.com\u002Fmax\u002Fchangelog\n[MAX Model Development]: \u002Fmax\u002Fdocs\u002Fdevelopment.md\n\n\u003C!-- News & Announcements links -->\n[public-com-meet-doc]: https:\u002F\u002Fmodul.ar\u002Fcommunity-meeting-doc\n[bentoml-github]: https:\u002F\u002Fgithub.com\u002Fbentoml\u002FBentoML\n[bentoml-joins-ama]: https:\u002F\u002Fforum.modular.com\u002Ft\u002Fmodular-has-acquired-bentoml-ask-us-anything\u002F2706\u002F1\n[bentoml-joins]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fbentoml-joins-modular\n[26.2]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodular-26-2-state-of-the-art-image-generation-and-upgraded-ai-coding-with-mojo\n[26.1]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodular-26-1-a-big-step-towards-more-programmable-and-portable-ai-infrastructure\n[mojo-1.0]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fthe-path-to-mojo-1-0\n[dec-meetup]: https:\u002F\u002Fwww.youtube.com\u002Flive\u002FWK5dVQ8vhbU?si=Fjde8j_50V4bwiAv\n[25.7]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodular-25-7-faster-inference-safer-gpu-programming-and-a-more-unified-developer-experience\n[250-funding]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodular-raises-250m-to-scale-ais-unified-compute-layer\n[pytorch-llvm]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fpytorch-and-llvm-in-2025-keeping-up-with-ai-innovation\n[25.6]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodular-25-6-unifying-the-latest-gpus-from-nvidia-amd-and-apple\n[25.5]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodular-platform-25-5\n[la-meetup]: https:\u002F\u002Flu.ma\u002Fmodular-aug-meetup\n[amd]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodular-x-amd-unleashing-ai-performance-on-amd-gpus\n[hack-weekend]: https:\u002F\u002Fwww.meetup.com\u002Fmodular-meetup-group\u002Fevents\u002F308311461\u002F\n[hackathon]: https:\u002F\u002Fwww.modular.com\u002Fblog\u002Fmodverse-48\n\n\u003C!-- Community & Events links -->\n[discord]: https:\u002F\u002Fdiscord.gg\u002Fmodular\n[forum]: https:\u002F\u002Fforum.modular.com\u002F\n[meetup-group]: https:\u002F\u002Fwww.meetup.com\u002Fmodular-meetup-group\u002F\n[youtube]: https:\u002F\u002Fwww.youtube.com\u002F@modularinc\n","Modular Platform 是一个统一的AI开发和部署平台，包括MAX和Mojo两个核心组件。它提供了一套完整的AI库和工具，能够加速模型服务并扩展生成式AI应用的部署规模，同时抽象了硬件复杂性，使用户无需修改代码即可在GPU和CPU上以行业领先的性能运行最流行的开源模型。该平台支持通过pip或conda安装，并允许用户快速启动与OpenAI兼容的服务端点。此外，还提供了Kubernetes兼容的Docker容器用于便捷部署，适用于需要高效管理和扩展AI模型的企业级应用场景。",2,"2026-06-11 03:23:29","top_topic"]