SpaceXAI releases Grok 4.7, which it says is better at verifying its own work and managing longer context, available for $2/1M input and $6/1M output tokens

Input token costs for this model are now half of what they were for comparable advanced models.

SpaceXAI has launched Grok 4.7, a new AI model designed for coding and knowledge work. The company claims this model is twice as fast and offers twice the output speed of its previous version, Grok 4.6, while maintaining the same price point. Grok 4.7 features improvements in verifying its own work and managing longer context windows, and it reportedly performs better on benchmarks like CursorBench 4.0 for coding tasks and Terminal-Bench 4.0. It also shows enhanced capabilities in professional knowledge work, legal tasks, and clinical reasoning, comparable to other leading models. SpaceXAI highlights its new safeguard stack, which it states is more robust against jailbreaking and refusals while maintaining utility for benign tasks. Grok 4.7 is available via API and other platforms at $2 per million input tokens and $6 per million output tokens.

The introduction of Grok 4.7 signifies a market push towards more capable and cost-effective AI models, particularly for complex coding and knowledge-intensive tasks. Its improved self-verification and context management suggest a move towards more reliable and sophisticated AI assistants that can handle intricate, multi-step processes with less human oversight. The competitive pricing, especially the $2/1M input token rate, directly challenges existing leaders, potentially driving down costs across the sector and making advanced AI more accessible for a wider range of applications and businesses.

Companies relying on AI for software development, legal analysis, or complex data processing can now expect better performance and potentially lower operational expenses. The enhanced safety features, balancing utility with robust refusal mechanisms for dangerous tasks, address growing concerns around AI misuse and could set a new standard for responsible AI deployment. The market should watch for how competitors respond to this price-performance benchmark and if similar cost reductions and capability upgrades become the new industry norm.