Modular, a Qualcomm company

AI shouldn’t be locked to one hardware vendor or chip. We’re building MAX to democratize AI compute, so developers and companies can build and deploy AI across hardware platforms without vendor lock-in. To achieve this vision, MAX is now source available. The graph compiler, runtime and hardware backends join the kernels, models, serving infrastructure and Mojo compiler + standard library we’ve already opened. The new MAX License balances two priorities: freedom to build, and a unified, sustainable ecosystem. Anyone can use, modify and ship commercial products with MAX, without licensing fees or royalties. Its few restrictions prevent fragmentation into incompatible versions, so everyone builds on a common foundation. The Modular alliance program extends that foundation, giving hardware vendors and partners a shared path to build on MAX together: https://lnkd.in/gG4VjyFA Also in 26.7: eager mode for faster custom op debugging, decision model support, including Decider 2B, and Mojo 1.2, the next step in Mojo’s growth into a systems programming language, with pattern matching. Everything in 26.7: https://lnkd.in/gaXj5g3T More on the new MAX License and our alliance program: https://lnkd.in/g98NQg9t

Companies are spending 4x more than they need on tokens. Amazingly accurate and cost efficient models are dropping every day. Modular has the latest one - GLM 5.3. Try it for free today, then we'll help your team scale it up, to bring your inference costs down.

GLM 5.3 is here - on Modular Cloud! It just dropped this week, so come try it out for free. This is Z.ai's most capable model for agentic coding and cyber defense is now available to download, run, and customize.

When NVIDIA supply tightens, being locked to one vendor is a liability. GLM 5.3 runs on NVIDIA and AMD on Modular Cloud with zero code changes. Test it out today. Start free.

Your users want GLM 5.3. It's already tuned and serving on Modular Cloud, behind an OpenAI-compatible API. Swap your base URL, keep your code, ship today. Signing up is free — send your first request in minutes.

GLM 5.3 on Modular Cloud - TODAY! With Modular you can run AI across GPUs, CPUs, and ASIC - engineered for the most demanding inference workloads. Sign up for Modular Cloud for free today, test GLM 5.3 and other open source models.