This is huge! State-of-the-art Llama3 🦙 perf on Mac? Look no further, try MAX today! But wait!✋MAX isn’t just a model specific disposable framework; it’s a rearchitecting of AI infra. built from the ground up by the best AI compiler and infra team in the world. MAX⚡️🤝Mojo🔥=🏆

New blog post! I dive deep into what's new in the latest Mojo🔥 24.4 with lots of code examples. Improved Mojo collections, write math friendly code with new traits, new file I/O features, perf. improvements and lots more! Jupyter Notebook included! modular.com/blog/whats-new…

An end of an era: anandtech.com/show/21542/end…

🤯 what an achievement, the sheer complexity, trajectory planning, control design, all the modeling and simulation to get this in 1 attempt. AI is cool, but this is gives me Saturn V and moon landing levels of excitement 🚀

MLX was open sourced exactly one year ago 🥳. It's now the second most starred and forked open source project from Apple -- after the Swift programming language.

Latest mlx + mlx-lm have much faster prompt processing speeds for MoEs. Thanks to some magic from Angelos Katharopoulos pip install -U mlx-lm Deep Seek v3 → up to 2x faster Mixtral → up to 3.5x faster Llama 4 → up to 2x faster

Exciting MLX updates at #WWDC25 new webpage and 2 sessions for Python and Swift devs! New MLX webpage! mlx-framework.org Getting started with MLX by Awni Hannun developer.apple.com/videos/play/ww… Explore LLMs on Apple silicon by Angelos Katharopoulos developer.apple.com/videos/play/ww…

We have two awesome new videos on MLX at #WWDC25 this year. - Learn all about MLX. - Learn all about running LLMs locally with MLX. Angelos Katharopoulos, Shashank Prasanna, myself, and others worked super hard to make these. Check them out and hope you find them useful!

We released accompanying code for our #WWDC25 sessions. We have Jupyter notebooks for Python developers and Xcode projects for Swift developers! github.com/ml-explore/mlx… Question about MLX? Register for 1:1 online labs on Wednesday and Friday: developer.apple.com/wwdc25/session…

A lot of care went into this session by Angelos Katharopoulos. You can do so much with a single line command using MLXLM CLI. (1) quantizing: `mlx_lm.convert` (2) finetuning: `mlx_lm.lora` (3) fusing and uploading to Hugging Face : `mlx_lm.fuse` Entire notebook: : github.com/ml-explore/mlx…

From Foundation Models framework to open-source MLX, CoreML, Vision, Speech and more! Lot of exciting AI & ML announcements @WWDC25 this week! Questions? Join us tomorrow for a live online Q&A and deep dive at the ML & AI group lab: developer.apple.com/wwdc25/labs/45…

We heard developer feedback on Monday that the MLX Swift LLM API is hard to get started with. So we went to work. We made a new and improved streamlined API. Now 3 lines to load an LLM or VLM and start a chat session in a Swift project. Happy last day of #WWDC25 !

🚀 Excited to launch Qwen3 models in MLX format today! Now available in 4 quantization levels: 4bit, 6bit, 8bit, and BF16 — Optimized for MLX framework. 👉 Try it now! Huggingface：huggingface.co/collections/Qw… ModelScope： modelscope.cn/collections/Qw…

MLX 🤝 GPUs! MLX with NVIDIA AI Developer CUDA backend!

If you missed the WWDC25 recap broadcast, video is now on Youtube! I cover: - On-device LLM using the Foundation Models framework - Custom models using CoreML - Latest research models using open-source MLX and more! Enjoy! youtube.com/watch?v=9lO3ds…

🤯

Generate live captions right in your browser using FastVLM and Metal support for WebGPU! Test it out on huggingface: huggingface.co/spaces/apple/f… (If you are on an older iOS version, enable WebGPU support in safari > advanced > feature flags)

Always a good day with a new MLX release drops🎊: MXFP4 precision support and NCCL for MLX distributed on NVIDIA GPUs😱

Come join us LIVE and build an on-device LLM-powered app at 10 AM PT Today!! 🚀 2-hour code-along with Foundation Models framework, incl.: ✅ Starter project & guide ✅ Live expert Q&A ✅ Guided generation, streaming, tool calling, perf. optim. & more! youtube.com/watch?v=A8X6hN…