Ox Alpha is now the biggest model launch in OpenRouter history.
In its first 3 full days after launch, Ox Alpha processed 11.6 trillion tokens. 2.6× more than the next-biggest launch on OpenRouter.
Ox Alpha is on track to hit nearly 6 trillion tokens today.
Try it now via ori in your coding agents:
$ ori [your favorite harness] --model stealth/ox-alpha
OpenAI is bringing back the 5-hour limit for Plus accounts across ChatGPT Work and Codex starting tomorrow.
The 5-hour limit will remain disabled for Pro $100 and Pro $200 subscriptions.
Tomorrow we will bring back the 5h limit for Plus accounts across ChatGPT Work and Codex. I had mentioned this a while ago, but then postponed it.
This is necessary as (a) the 5h limit allows us to smoothen the load on our compute, allowing to keep the plan generous in terms of
Browser Use is now available in Grok Build.
The integration lets users ask Grok to complete browser-based tasks, as well as test websites and fix bugs.
Grok Build can work with:
• The user’s own browser
• A browser hosted in the cloud
To install Browser Use in Grok Build:
Browser Use is now available in @SpaceXAI's Grok Build!
Ask Grok to complete tasks for you, or have it test your websites and fix bugs.
Grok Build can work in your own browser or one hosted in the cloud.
Cline gave Ox Alpha and Fable 5 the same real bug.
Both fixed it.
But Fable said “I found the root cause” 7 TIMES before editing.
Ox said it once and fixed the thing.
~3× fewer output tokens for the same job.
We benchmarked Ox Alpha vs Fable on a real bug from the Cline repo. Both fixed it correctly.
But we found that Ox used much fewer thinking tokens. Most reasoning models loop and re-derive the same conclusion over and over before acting (Fable said "I found the root cause" 7
