xAI Releases Grok 4.7, Focused on Coding and Knowledge Work
Introducing Grok 4.7 SpaceXSI's most powerful model for coding and knowledge work. Twice as fast, at half the price of comparable models. Sep 21, 2026
AI BriefingxAI releases Grok 4.7, positioning it as its strongest model for coding and knowledge work. It uses a larger base model than Grok 4.6 and undergoes longer reinforcement learning training. Pricing is $2 per million input tokens and $6 per million output tokens, with a faster version that doubles both output speed and price; the company says its price-performance ratio is frontier-level on CursorBench 4.0, while HackerBench v0.3 allowed only 3.3% of dangerous dual-use prompts.
Another 10 sources reported this
- xAI Releases Grok 4.7: Larger Base Model, Maintaining the Same $2/$6 Pricing as Grok 4.6MarkTechPost · September 22
- xAI releases Grok 4.7Elon Musk · September 22
- SpaceXAI releases Grok 4.7, focused on coding and knowledge processing, starting at $2 per million input tokensITHome · Artificial Intelligence · September 22
- Grok 4.7 Released: 19.6% Lead on the Harvey Legal Benchmark, Nearly Double the Terminal-Bench 4.0 Score, with Unchanged PricingRohan Paul · September 22
- Grok 4.7 model card released, with improvements over 4.6 on multiple benchmarks and unchanged pricingEric Zakariasson · September 22
- xAI releases Grok 4.7, which is inexpensive but significantly trails Claude and GPT-6 on benchmarksThe Decoder: AI News · September 22
- xAI releases Grok 4.7, focused on coding and long-running tasksHacker News: Popular AI Posts · September 22
- Grok 4.7 released; the author reposts claims that it delivers significant improvements over Grok 4.6 with no change in price or speedLauren Tan · September 22
- Grok 4.7 balances intelligence, speed, and low costElon Musk · September 22
- Grok 4.7 released with improved capabilities at unchanged speed and pricingMichael Truell · September 22
Source: xAI: News · x.ai
Event follow-up3
View all event coverage- Kazzak’s Hands-On Comparison of Grok 4.7 and Xiaomi MiMo V2.6: The Latter Is the Definitive Answer to the Impossible TriangleWeChat public account: Digital Life Kazik · September 22
- Artificial Analysis evaluates Grok 4.7: its Intelligence Index score of 46 places it in the top four, while its coding agent rises to fourth placeArtificial Analysis full article · September 22
- xAI Releases Grok 4.7, Focused on Coding and Knowledge WorkxAI: News · September 21