intenAi🦋

Really cool work from the team reimagining the mouse pointer to be intelligent! Try the prototype in @GoogleAIStudio it's pretty magical.

Imagine every pixel on your screen, streamed live directly from a model. No HTML, no layout engine, no code. Just exactly what you want to see. @eddiejiao_obj, @drewocarr and I built a prototype to see how this could actually work, and set out to make it real. We're calling it Flipbook. (1/5)

Last week, we released a preview of memories in Codex. Today, we’re expanding the experiment with Chronicle, which improves memories using recent screen context. Now, Codex can help with what you’ve been working on without you restating context.

Codex is open source, enabling anyone to build awesome applications on top of it:

Build your own assistant with @NVIDIAAI. ❤️

codex is becoming a full agentic IDE

AI will create new kinds of work, new industries, and new ways to build. Last week at Stanford, our CEO Jensen Huang joined Rep. Ro Khanna for a conversation on AI, jobs and the long-term opportunity ahead.

NVIDIA just dropped Lyra 2.0. This AI can turn one image into an explorable 3D world. Export to 3D Gaussians, meshes, and physics engines. Model Card + Project👇

Today we launched Gemini 3.1 Flash TTS, our most expressive and controllable text-to-speech model yet. This launch [excitement] includes audio tags! 🗣🏷 Audio tags [explanatory] are a seamless way to guide vocal style, pace, and delivery using natural language commands embedded directly in your text. Want a different tempo or tone? [amazement] Just tag the audio to steer the AI-speech output! The model supports 70+ languages (24 of which are high-quality evaluated languages, including: Japanese, Hindi, and Arabic). Watch the audio tags in action in the demo below ↓

Our most expressive and steerable TTS model yet! Designed to give builders granular control over AI-generated speech, Gemini 3.1 Flash TTS is really fun to play with! Available in preview today - for devs via the Gemini API & @GoogleAIStudio + for enterprises on Vertex AI

Today, we released Lyra 2.0, a framework for generating persistent, explorable 3D worlds at scale, from NVIDIA Research. Generating large-scale, complex environments is difficult for AI models. Current models often “forget” what spaces look like and lose track of movement over time, causing objects to shift, blur, or appear inconsistent. This prevents them from creating the reliable 3D environments required for downstream simulations. Lyra 2.0 solves these issues by: ✅ Maintaining per-frame 3D geometry to retrieve past frames and establish spatial correspondences ✅ Using self-augmented training to correct its own temporal drifting. Lyra 2.0 turns an image into a 3D world you can walk through, look back, and drop a robot into for real-time rendering, simulation, and immersive applications. ➡️ Learn more: research.nvidia.com/labs/sil/proje… 📄 Read the paper: arxiv.org/abs/2604.13036

Introducing Gemini on Mac. It’s the first time we’re bringing the @Geminiapp to desktop. The team built this initial release with @Antigravity, and it went from an idea to a native Swift app prototype in a few days. More features on the way!

Great to see our collaboration w/ @BostonDynamics unlocking new capabilities! Gemini Robotics-ER 1.6 enables robots like Spot to read complex industrial gauges autonomously. Exciting step toward robots that can understand & operate usefully in the physical world

Ollama is now updated to run the fastest on Apple silicon, powered by MLX, Apple's machine learning framework. This change unlocks much faster performance to accelerate demanding work on macOS: - Personal assistants like OpenClaw - Coding agents like Claude Code, OpenCode, or Codex

⚡️ Introducing Mistral Moderation 2, our next-generation moderation model. It introduces new categories and builds on the strengths of the previous version. - Enhanced performance - 128k context length (up from 8k) - Free to use

Today, we’re introducing Forge, a system for enterprises to build frontier-grade AI models grounded in their proprietary knowledge. 🌎 Forge bridges the gap between generic AI and enterprise-specific needs. Instead of relying on broad, public data, organizations can train models that understand their internal context embedded within systems, workflows, and policies, aligning AI with their unique operations. We have already partnered with world-leading organizations, like ASML, DSO National Laboratories Singapore, Ericsson, European Space Agency, Home Team Science and Technology Agency (HTX) Singapore and Reply to train models on the proprietary data that powers their most complex systems and future-defining technologies.

🔥 Meet Mistral Small 4: One model to do it all. ⚡ 128 experts, 119B total parameters, 256k context window ⚡ Configurable Reasoning ⚡ Apache 2.0 ⚡ 40% faster, 3x more throughput Our first model to unify the capabilities of our flagship models into a single, versatile model.

Ready to deploy AI agents? NVIDIA NemoClaw simplifies running @openclaw always-on assistants with a single command. 🦞 Deploy claws more safely ✨ Run any coding agent 🌍 Deploy anywhere Try now with a free NVIDIA Brev Launchable 🔗 nvidia.com/nemoclaw

I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend. It's basically nanochat LLM training core stripped down to a single-GPU, one file version of ~630 lines of code, then: - the human iterates on the prompt (.md) - the AI agent iterates on the training code (.py) The goal is to engineer your agents to make the fastest research progress indefinitely and without any of your own involvement. In the image, every dot is a complete LLM training run that lasts exactly 5 minutes. The agent works in an autonomous loop on a git feature branch and accumulates git commits to the training script as it finds better settings (of lower validation loss by the end) of the neural network architecture, the optimizer, all the hyperparameters, etc. You can imagine comparing the research progress of different prompts, different agents, etc. github.com/karpathy/autor… Part code, part sci-fi, and a pinch of psychosis :)

nanochat now trains GPT-2 capability model in just 2 hours on a single 8XH100 node (down from ~3 hours 1 month ago). Getting a lot closer to ~interactive! A bunch of tuning and features (fp8) went in but the biggest difference was a switch of the dataset from FineWeb-edu to NVIDIA ClimbMix (nice work NVIDIA!). I had tried Olmo, FineWeb, DCLM which all led to regressions, ClimbMix worked really well out of the box (to the point that I am slightly suspicious about about goodharting, though reading the paper it seems ~ok). In other news, after trying a few approaches for how to set things up, I now have AI Agents iterating on nanochat automatically, so I'll just leave this running for a while, go relax a bit and enjoy the feeling of post-agi :). Visualized here as an example: 110 changes made over the last ~12 hours, bringing the validation loss so far from 0.862415 down to 0.858039 for a d12 model, at no cost to wall clock time. The agent works on a feature branch, tries out ideas, merges them when they work and iterates. Amusingly, over the last ~2 weeks I almost feel like I've iterated more on the "meta-setup" where I optimize and tune the agent flows even more than the nanochat repo directly.