Darius Baruo
Sep 21, 2026 21:13
SpaceXAI unveils Grok 4.7, its most powerful AI for coding and knowledge tasks, offering top benchmarks at competitive pricing.
SpaceXAI has officially launched Grok 4.7, its most advanced AI model for coding and professional knowledge work. Announced on September 21, 2026, Grok 4.7 delivers twice the speed and half the cost compared to similar models, making it a strong contender in the competitive AI landscape.
The model introduces several key improvements over its predecessor, Grok 4.6. It was trained on a larger base model and subjected to longer reinforcement learning with a focus on complex, multi-hour tasks. Notably, Grok 4.7 demonstrates enhanced self-verification capabilities and better handling of extended contexts, making it ideal for professional settings like software engineering and legal analysis. On CursorBench 4.0, a benchmark for extended coding tasks, Grok 4.7 outperforms Grok 4.6 with a 46.3% score, compared to 40.4% for the previous version.
Pricing for Grok 4.7 starts at $2 per million input tokens and $6 per million output tokens, aligning it with Grok 4.6 but significantly undercutting competitors like GPT-5.6 Sol ($4 and $20, respectively) and Fable 5.1 ($10 and $50). SpaceXAI also offers a faster variant at double the speed for double the price, providing flexibility for developers and businesses with varying needs.
Beyond its technical upgrades, Grok 4.7 also prioritizes safety and cybersecurity. It features a newly designed safeguard stack, achieving top scores in benchmarks like HackerBench v0.3 and LatchBio’s biosafety test. The model has been particularly effective in refusing malicious prompts while maintaining usability for legitimate tasks, making it a trusted tool for cybersecurity professionals. Select industry partners now have access to its specialized red-team capabilities for defense research.
Artificial Analysis has rated Grok 4.7 with an Intelligence Index score of 46, placing SpaceXAI among the top AI labs globally. While its coding-agent performance is competitive with leading models from Anthropic and OpenAI, Grok 4.7 distinguishes itself with its cost-efficiency and focus on professional use cases rather than general intelligence.
Grok 4.7 is available immediately through the xAI API, as well as platforms like Cursor and Grok Build. Interested users can also try it for free at x.ai/build.
For developers and enterprises, Grok 4.7 represents a significant step forward in AI performance and accessibility. Its combination of speed, accuracy, and affordability could make it a preferred choice for coding tasks and knowledge work in the months to come.
Image source: Shutterstock
Credit: Source link
