xAI launches Grok 4.7 for coding and knowledge work at $2 input and $6 output tokens
xAI has launched Grok 4.7, its newest model for coding and knowledge work, priced at $2 per million input tokens and $6 per million output tokens. The model uses a larger base model than Grok 4.6, features longer reinforcement learning on harder tasks, and includes improved safeguards and self-verification capabilities. It is available via Grok API, Cursor, Grok Build, and third-party platforms.
Key points
- xAI launched Grok 4.7 for coding and knowledge work at $2 per million input and $6 per million output tokens.
- Grok 4.7 uses a larger base model and longer reinforcement learning run than Grok 4.6.
- Benchmark results differ: unite.ai reports 46.3% on CursorBench 4.0; the-decoder.com reports 46 on AA Index and 26% on Terminal-Bench 4.0.
While all sources agree on the core launch details, pricing, and model improvements, they differ in benchmark reporting: unite.ai reports Grok 4.7 scoring 46.3% on CursorBench 4.0 and 71.0% on DeepSWE v1.1, while the-decoder.com cites a score of 46 on the Artificial Analysis Intelligence Index v4.3.2 and 26% on Terminal-Bench 4.0, noting a gap to Claude and GPT-6. aws.amazon.com does not discuss Grok 4.7, focusing instead on Grok 4.6’s availability in Amazon Bedrock.
Model pages: Claude Fable 5.1 → · GPT-6 →
The story so far
2 episodes →- xAI launches Grok 4.7 for coding and knowledge work at $2 input and $6 output tokensthis story
Introducing Grok 4.7
x.ai · 21 September 2026
Loading the full article…
This text was published by x.ai and written by xAI. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
5sources- Reddit discussionreddit.com
- SpaceXAI Releases Grok 4.7: A Larger Base Model at the Same $2/$6 Price as Grok 4.6Press · MarkTechPost ·
- xAI’s Grok 4.6 is now available in Amazon BedrockPrimary source · AWS Machine Learning Blog ·
- xAI launches Grok 4.7 at bargain prices, but benchmarks reveal a wide gap to Claude and GPT-6Press · The Decoder ·
- SpaceXAI Releases Grok 4.7 for Coding and Knowledge WorkPress · Unite.AI ·
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Google launches Gemini 4 Argon model globally for Pro and Ultra subscribers with introductory API pricing · 32 src
- Meta releases Muse Spark Open, an 816B-parameter open-weight multimodal model with 42B active parameters and 1M-token context · 20 src
- Microsoft AI chief says in-house model beats OpenAI's using quarter of compute · 1 src
- NVIDIA releases Nemotron 3.5 Super VL and Bolt edge models for enterprise agents · 8 src
- Black Forest Labs releases FLUX 3 Image Klein, a 5.6B-Parameter Open-Weights model for text-to-image generation · 3 src
Comments
via GitHub Discussions