🚨Grok 4.7 is out and is looking ridiculous on CursorBench.
It gets 99% of Opus 5 Max’s score at ~50% of the cost:
46.3% @ $6.01/task vs 46.6% @ $11.95.
It also beats GPT-5.6 Sol Max while costing 27% less.
Frontier labs like @OpenAI and @AnthropicAI are likely already training their own versions of Jev.
There are tons of real world use cases where Jev can outperform standard LLMs and I’m honestly surprised we didn’t see something like this sooner.
This feels like a very obvious next step for AI.
Breaking: Browser Use + Jev = Ultrafast ⚡
Findings flights took 7s and cost only $0.0039 🤯
> new action space every step
> DOM state space
> small LLM fallback to type
(this video is at 1x speed btw)
Built a tiny open source browser agent. try it below ↓
Cancelled my grok bot sub today, even tho the product, ui/ux is very nice, I would rather use that subscription amount for another ChatGPT/Claude account.
If you are a grok bot user how’s your experience after 1 month.
Grok 4.7 release timeline:
Sept 1: “10 days”
Sept 11: “Few more days”
Sept 16: 👀
We are no longer waiting for a model release.
We are participating in a psychological experiment. 😂
Who else is frantically checking X every hour for the drop? 👀
Does posting cringy stuff like “gpt astra for sales” actually work?
Has anyone found success in this? I am actually curious if it’s a good strategy or overused now?