xAI Launches Grok 4.7 With Coding Gains
xAI Launches Grok 4.7 With Coding Gains
Anonymous
Key Points
1
The launch followed at least five timeline revisions by Elon Musk since late July, including promises of release in “four weeks,” “a few weeks,” “3 to 4 weeks,” “10 days” and, most recently, “a few more days to cook.”
2
Grok 4.7 offers four reasoning settings—low, medium, high and xhigh—alongside its 500,000-token context window.
3
On the GDPval professional-work benchmark, Grok 4.7 scored 1,695 Elo, below Claude Fable 5.1’s 1,735, indicating that it remained competitive but did not lead the field on every evaluation.
4
xAI reported specific safety-test results: Grok 4.7 allowed 3.3% of risky dual-use prompts on HackerBench 0.3 and scored 62.4% on a LatchBio biosafety benchmark.
5
The release was available without a waitlist through the Grok app as well as Cursor, Grok Build and the xAI API.
xAI launched Grok 4.7 on September 21, positioning it as a major coding upgrade after several delays. The model uses 2.1 trillion parameters—40% more than its predecessor—and includes a 500,000-token context window, making it suited for extended reasoning and professional work Encrypted TechI.
The release followed at least five timeline revisions by Elon Musk since late July, with promises ranging from 'four weeks' to 'a few more days to cook' Encrypted. Pricing stays at $2 per million input tokens and $6 per million output tokens, with a faster version costing double TechI.
Sentiment
Mostly PositiveNEGATIVENEUTRALPOSITIVE
Publishers
15
Articles
4
Reach
19
A
AnonymousAuthor
