{"slug": "modular-modular-26-7-opening-up-max", "title": "Modular: Modular 26.7: Opening up MAX", "summary": "Modular released version 26.7 of its MAX platform on October 22, 2026, making the MAX graph compiler, MAX runtime, and device extensibility hooks fully source-available under a new MAX License that permits use, modification, extension, and commercial products with no fees or royalties. The release also adds full eager mode support, including for custom ops, and decision model support, with source for MAX plugins for Qualcomm Dragonfly and Google TPU planned for release in the coming months. Modular said the permissive license differs from Apache 2.0 only on fragmentation, which it addresses through its alliance program for hardware vendors and partners.", "body_md": "Inference Products\n\nShared Endpoints\n\nAccess frontier models via an API\n\nDedicated Endpoints\n\nMission critical reliability\n\nCustom models\n\nYour model, peak performance\n\nDeployment Options\n\nOur Cloud\n\nFully managed, pay by usage\n\nYour Cloud\n\nModular stack in your VPC\n\nPricing\n\nFlexible plans for every team\n\nModels\n\nFLUX.2 Klein 4B\n\nGLM 5.3 Flash\n\nMiniMax M3\n\nGLM 5.3\n\nView All\n\nText to audio\n\nTurn text into natural speech\n\nImage generation\n\nGenerate images from text prompts\n\nCode generation\n\nGenerate production-ready code\n\nVideo generation\n\nGenerate video from text + image\n\nAgentic\n\nDeploy AI agents anywhere\n\nCustom Models\n\nKernel-level model control\n\nCase Studies\n\nProven results from real customers\n\nModels\n\nFLUX.2 Klein 4B\n\nGLM 5.3 Flash\n\nMiniMax M3\n\nGLM 5.3\n\nView All\n\nModel Library\n\nLatest supported open models\n\nLLM Inference Handbook\n\nHow inference works, from basics to production\n\nMAX LLM Book\n\nBuild a modern LLM from the ground up\n\nMojo Quest\n\nLearn Mojo through guided challenges\n\nMojo GPU Puzzles\n\nHands-on GPU programming, puzzle by puzzle\n\nMAX\n\nGenAI native modeling & serving\n\nMojo\n\nThe systems language for the AI era\n\nModular alliance program\n\nLLVM, but for AI\n\nSelf-Hosted\n\nMAX+Mojo self-hosted by you\n\nCommunity\n\nBuild the future of AI together\n\nAgentic Skills\n\nOfficial AI agent skills from Modular\n\nModular Cloud\n\nDeploy GenAI models, our cloud or yours\n\nMAX\n\nGenAI native modeling & serving\n\nMojo\n\nThe systems language for the AI era\n\nAbout\n\nBuild AI for anyone, anywhere.\n\nCareers\n\n👋 We’re currently hiring!\n\nCulture\n\nWhat we believe\n\nContact Us\n\nRequest a demo\n\nOctober 22, 2026\n\nModular Team\n\nProduct\n\nThe headline of Modular 26.7 is one we've wanted to write for a long time: this release opens the MAX graph compiler, MAX runtime, and device extensibility hooks for everyone. This means that MAX is now fully source-available, and can be built and modified by hardware innovators, researchers, and everyone else who loves making accelerators go \"brr.\" This release also brings an eager mode that gets you debugging custom ops in a fraction of the time, decision model support, and the next step in Mojo's growth into a systems programming language.\n\nAt our recent ModCon conference, we committed to opening up MAX, and we're delivering on that promise.\n\nThe full source for MAX is now available: hardware backends, the graph compiler, the runtime, and more. Our new MAX License is intentionally permissive: you can use, modify, extend, and build commercial products with MAX, without licensing fees or royalties. If you've wanted to bring up MAX on your own accelerator, this is where to start, and you can see the existing implementation for NVIDIA, AMD, Apple Silicon, and CPUs as a reference point. We plan to release source for the MAX plugins for other chips, including Qualcomm Dragonfly and Google TPU, in the coming months.\n\nThe MAX License allows nearly everything Apache 2.0 does: use, modify, extend, and build commercial products with MAX, with no fees or royalties. The difference comes down to fragmentation. We’ve spent decades maintaining major open source projects and have seen the challenges they face. MAX aims to unify the programming model across hardware, and incompatible platform-specific versions would undermine that goal. The Modular alliance program is how we solve this, giving hardware vendors and partners a shared path to build on MAX without splintering the ecosystem. As with Mojo, we expect our approach to evolve as the technology matures and we hear from the community.\n\nThere's far more in the MAX source code than we can cover here, so we wrote a separate post on what's included and what you can build with it.\n\nMAX has long had an advanced graph compiler that performs automatic fusion and other optimizations for peak performance. But during development, fast iteration often matters more than peak performance. MAX now has full eager mode support, including your custom ops, and it shares a single code path with the graph compiler, so the code you run in eager mode is the same as what you compile and ship.\n\nDeclare an op once, call it directly on a Tensor, then reuse it unchanged inside a compiled Module. That makes the development loop much faster: on a slice of Gemma 4 31B, eager mode reaches the first token in about 30 seconds, compared to about 110 seconds compiled.\n\nHere’s an example of how eager mode works with custom ops:\n\n```\nn, w, c = custom.Symbols(\"n\", \"w\", \"c\")\ndownsample = custom.declare(\n    \"downsample\",\n    inputs={\"x\": custom.TemplateType(DType.float32, [n, w, c])},\n    outputs=[custom.TemplateType(DType.float32, [n, w // 2, c])],\n    custom_extensions=[\"path/to/kernels\"],\n)\n\n# Eager: runs immediately.\ny = downsample(Tensor.ones([2, 8, 3]))        # shape (2, 4, 3)\n\n# Compiled: same op inside a Module.\nclass Encoder(Module):\n    def forward(self, x: Tensor) -> Tensor:\n        return downsample(x) * 2.0\n\nmodel = Encoder().compile(TensorType(DType.float32, [\"batch\", 8, 3], device=Encoder().device))\ny = model(Tensor.ones([2, 8, 3]))             # same op, now compiled\n```\n\nThere’s been a lot of recent excitement about the new modality of System One/decision models, due to their ability to enable zero-shot classification in a very inexpensive manner. MAX has added initial support for these decision models, including Decider 2B, through the new v1/decisions and v1/systemone API endpoints.\n\nMAX has also added support for MiMo-V2.6-Flash, GLM-5.3-Flash, and Muse-Glimmer, with KDA and sparse MLA.\n\nIn addition to new capabilities, we continue to improve existing model families. Several ops that used to round-trip through the CPU now run entirely on the GPU, which gives a 2.6x prefill speedup for Qwen and GLM models.\n\nWith 1.0, Mojo proved itself as a language for high-performance code across heterogeneous hardware. Now we're leaning into the next phase of the Mojo roadmap: Mojo as a systems programming language.\n\nMojo 1.2's flagship new feature is pattern matching. The match statement lets you replace long chains of conditionals with clear, structured cases, and the compiler can check that you've handled every one of them.\n\nmatch\n\n``` php\ndef classify_point(p: Tuple[Int, Int]) -> String:\n    match p:\n        case 0, 0:\n            return \"origin\"\n        case x, 0 if x < 0:\n            return String(\"negative x-axis:\", x)\n        case x, 0:\n            return String(\"x-axis:\", x)\n        case 0, y:\n            return String(\"y-axis:\", y)\n        case x, y:\n            return String(\"point:\", x, \",\", y)\n```\n\nAs a bonus, it also works at compile-time:\n\n``` php\ndef classify_type[T: AnyType]() -> String:\n    comptime match T:\n        case == Int:\n            return \"integer\"\n        case == String:\n            return \"string\"\n        case _:\n            return \"other\"\n```\n\nStructs can now explicitly opt out of trait conformance (for example, by declaring themselves not Movable), and where clauses accept an else branch so you can state exactly why a constraint failed. Both make generic code easier to read and easier to debug.\n\nnot Movable\n\nwhere\n\nelse\n\npip install --upgrade max\n\nWith the whole stack open, we can’t wait to see what you build. Share it on the Modular forum.\n\nMAX is now source available and open to everyone\n\nOctober 22, 2026\n\nModular 26.5: Mojo 1.0 is here!\n\nAugust 11, 2026\n\nModular 26.4: SOTA MoE Serving, Model Bringup via Agent Skills, Mojo 1.0 Beta 2 and More\n\nJune 18, 2026\n\nBuild the future of AI with Modular\n\nSign up today\n\nSignup to our Cloud Platform today to get started easily.\n\nBrowse open models\n\nBrowse our model catalog, or deploy your own custom model\n\nGet all our latest news, announcements and updates delivered directly to your inbox. Unsubscribe at anytime.\n\n⚠️ This form requires JavaScript to function. Please enable JavaScript in your browser to continue.\n\nYour data is handled in accordance with Modular's Privacy Policy.\n\nThanks for signing up to our newsletter! 🚀\n\nThank you,\n\nModular Sales Team", "url": "https://wpnews.pro/news/modular-modular-26-7-opening-up-max", "canonical_source": "https://www.modular.com/blog/26-7", "published_at": "2026-10-22 00:00:00+00:00", "updated_at": "2026-10-22 18:14:23.535356+00:00", "lang": "en", "topics": ["ai-infrastructure", "developer-tools", "mlops", "ai-tools"], "entities": ["Modular", "MAX", "Mojo", "ModCon", "NVIDIA", "AMD", "Apple Silicon", "Qualcomm Dragonfly"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/modular-modular-26-7-opening-up-max", "markdown": "https://wpnews.pro/news/modular-modular-26-7-opening-up-max.md", "text": "https://wpnews.pro/news/modular-modular-26-7-opening-up-max.txt", "jsonld": "https://wpnews.pro/news/modular-modular-26-7-opening-up-max.jsonld"}}