Yuhuai (Tony) Wu

Solve math and understand the universe

I took Grok 4 for a spin this weekend to build this game prototype. I used SuperGrok Chat to generate the initial game prototype and then brought it over to Cursor to continue coding with Grok 4 MAX. Grok 4 in Cursor is like a no-nonsense agent. Doesn't speak much, but…

I'm hiring for our AI safety team at xAI! We urgently need strong engineers/researchers to work across all stages of the the frontier AI development cycle: data, training, evals, and product 1. job-boards.greenhouse.io/xai/jobs/47992… 2. job-boards.greenhouse.io/xai/jobs/47992…

we did some cool work to get a good model, sir - x.ai/news/grok-4. ps: looking for researchers and engineers interested reasoning post-training. reach out if you are interested.

Grok 4's adoption slope on Poe is one of the steepest we've seen among reasoning models in the days following launch. It only took three days to beat the previous week-one record holder, o3. (1/2)

The benchmarks are overwhelmingly positive, but here's what the Cline community is saying about Grok 4 after a few days: Pattern we're seeing: Cline users are treating Grok 4 as a planning specialist. "The most insanely robust plan I have ever seen" -- actual quote from our…

My thoughts on Grok 4 Heavy after 12hrs: Crazy good! “Create an animation of a crowd of people walking to form “Hello world, I am Grok” as camera changes to birds-eye.” And it 1-shotted the *entire* thing. No other model comes close. Watch the full clip.

Grok 4 is at the SOTA on long context up to 192k. Gemini 2.5 Pro still edges out on 192k but Grok 4 was more consistent overall. Very very impressed, it's a GREAT model.

building reasoning agents w/ @YuchenHe07 @qhwang3 was so fun, and the next paradigm will be even cooler -- agents will solve far harder problems far faster

From the 1st RL training using tools on a mini reasoning model at 16% HLE till now building the smartest agent w/ @qhwang3 @ericzelikman , more fun and breakthroughs to go! 🤖

Pretrain scale RL 😎

3 × 10 = 4

Looking forward to what Grok can do next - pushing frontiers of science and engineering, making new discoveries!

We built 200k-GPU clusters; We scaled up & curated higher-quality data; We scaled compute by 100x; We developed training & test-time recipes; We made everything RL native; We stabilized infrastructure and speeded up; That's how you turn RL into the pre-training scale. Yet I am…

xAI partners with @Polymarket to blend market predictions with X data and Grok’s analysis. Hardcore truth engine - see what shapes the world. This is just the start of our partnership with @Polymarket. More to come. 🚀

Grok can now generate charts. Currently works in browser, will come to more platforms in the coming days.

Please offer galaxy brain questions and answers. This is for Grok training. docs.google.com/forms/d/e/1FAI…

we made some fixes to your recommendation algorithm earlier this week this improves some longstanding issues with 1) your negative feedback not being taken into account 2) repeatedly seeing the same video 3) the SimCluster algorithm recommending irrelevant content early…

i'm all ears (and eyes) share your world with me

Voice mode upgrades are now live for all users on the Grok iOS app. Grok can: - See you and the world around you. - Search the web for you. - Speak in your chosen language. - Take on custom personas. - Resume conversations. Try it out on the Grok app: apps.apple.com/us/app/grok/id…

We have recently added multilingual and camera support to the voice mode. Give it a try!