⚡xAI released version 4.7 of its Grok AI model, claiming major performance gains.

"xAI releases Grok 4.7 with major performance gains" — Hacker News (best) · AI & Technology

22 Sept 2026

What it means

xAI has released 4.7, a new version of its large language model, claiming major improvements in coding and knowledge tasks. The model uses the same per-token pricing as its predecessor ($2 input / $6 output per million ) but reportedly consumes more than double the tokens, increasing effective cost per query. xAI also introduced a new and reported scores on (46.3%) and LatchBio biosafety (62.4%). The model is being integrated into GitHub Copilot and Microsoft Office products.

Accuracy & sourcesSourced

The claim is directly from xAI's official release and model card, making it a primary-source announcement. The benchmarks are self-reported and have not yet been independently verified by third parties, but the reporting from multiple outlets (VentureBeat, Forbes, DataCamp) consistently cites the same release details.

⚡xAI released version 4.7 of its Grok AI model, claiming major performance gains.

What it means

Accuracy & sourcesSourced

How new