Why aren't language learning apps like Duolingo investing more in local/on-device AI? Users get a better product (free "premium" features), companies get lower infrastructure costs, and it all works offline with no voice recordings sent to a server! Fable 5.2 cooked up this pop-up book game that runs LFM3-Omni from @LiquidAI locally in your browser w/ our custom WebGPU kernels. The same model does speech recognition, image understanding, conversation, and text-to-speech. Pretty wild. It also uses the new html-in-canvas API to make the entire experience more interactive and accessible!

ruNNtime v0.2.0 just got a huge performance boost 🚀 We integrated a lot of findings from @huggingface kernels library into ruNNtime, speeding inference up significantly. huge thanks for your help on this @nicodotdev @xenovacom 🤗

We've just made a breakthrough: the first true omni model running 100% locally in a browser... fully end-to-end, with no separate ASR or TTS models needed. With custom WebGPU kernels, we're able to run LFM3-Omni-2B at ~280 tok/s on my M4 Max, accepting visual, audio, or text and producing interleaved text and audio tokens. Just 90 ms to the first token and 140 ms to the first audio. The future of on-device multimodal AI agents is going to be wild.

EmbeddingGemma 2 maps text, images, video, audio, and code, running 100% locally on your device. At just 740M parameters, you can build powerful multimodal apps running entirely in your browser via WebGPU & Transformers.js. From multimodal search to cross-modal classification and clustering, local embeddings unlock endless possibilities.

The model is perfect for building multimodal search, retrieval, classification, and clustering applications. PS: Huge shout-out to @shadersweden for inspiring the liquid glass carousel UI! 🤗 Demo:

NEW: EmbeddingGemma 2 just released on Hugging Face. Project text, images, video, and audio into a single, unified embedding space. At just 740M parameters, it can even run 100% locally in your browser on WebGPU with 🤗 Transformers.js! Try it yourself 👇

You don't need Jev for good emoji search... just a tiny local model! Love this demo by @maxleiter, powered by a 33M-parameter embedding model running entirely in your browser (for free) with 🤗 Transformers.js!

Introducing: a coding agent (Pi) running entirely in your browser using a 2B model on WebGPU 🤯 MiniCPM5-2B + Pi, powered by Transformers.js + WebGPU + 4-bit ONNX weights. All previous attempt to create this failed but MiniCPM5 seems to make it usable. Available now on Hugging Face 👇

Two days ago we shipped @huggingface/kernels. Now here's the deep dive. Why we publish Jinja templates instead of WGSL files, how the browser compiles the fastest kernel for your GPU, and what that unlocks: → attention in 20 lines of JS, running on the GPU → 1M+ pixels animated by a single matmul, 400 fps vs 6 fps in plain JS → Fleet: benchmark your GPU, help us tune the kernels for every device out there Full video ↓

Super happy to share our intention to join forces with NVIDIA in a $12,930,300,000 acquisition 💛💚 10 years after starting Hugging Face, open-source AI is at an inflection point. Thanks to the community, we’ve shown that it can be a complement, and even an alternative, to closed-source APIs. But for it to happen at larger scale, it needs more compute, more support, more collaboration and more visibility. That’s why we went to talk to Jensen, who offered to do exactly that with us. In addition to doubling down on NVIDIA’s massive contributions to open-source AI (I called them the “King of American open-source AI” earlier this year), they’ve committed to strongly supporting Hugging Face and our mission while keeping the platform open, independent and compute agnostic. The founders and the team are all staying to keep pushing this mission forward. Together, we think we can make open source the default way to build AI, with the goal of empowering 100 million AI builders to own their intelligence rather than rent it. Excited about the next 10 years! 🤗🤗🤗

Introducing Fleet: GPU benchmarking entirely in your browser. Run WebGPU compute kernels drawn from real AI workloads directly on your hardware and earn a personalized card built for your device. Rate your GPU. Join the Fleet.

Superwhisper just released S1-mini, a 0.6B LLM designed to turn raw speech-to-text transcripts into clean written text. It can even run 100% locally in your browser on WebGPU, thanks to 🤗 Transformers.js! Try it out yourself! 👇

It's official: Qwen3.8-27B just scored 52 on the @ArtificialAnlys Intelligence Index. We now have an open-weight model that matches GPT-5.6 Luna (max) AND can run locally... even in your browser with custom WebGPU kernels! What a time to be alive! 🤯

Bonsai 27B just changed the local LLM game forever. 1-bit quantization shrinks it from 54GB to just 3.8GB (-93%), while retaining 90% of its intelligence. That's insane. With custom WebGPU kernels written by Fable 5 and GPT 5.6 Sol, the model now runs locally in your browser!

While we eagerly await Fable 5's return, our agentic WebGPU kernel optimization framework kept running. Opus 4.8 picked up where Fable left off, pushing Liquid AI's new LFM2.5 230M to an unbelievable 1,400 tok/s... running locally in your browser. Don't blink or you'll miss it.

I think Reachy is the one who needs chess lessons… 😅 Robotics meets WebAI: Gemma 4 running fully offline on WebGPU with Transformers.js, controlling Reachy Mini over WebSerial. No internet, just a browser and a USB-C cable. What should Reachy play next?