xAI released the Grok 4.7 language model

Company xAI introduced a new version of its language model Grok 4.7, focused on improving programming skills and working with knowledge bases, the update is already available for integration via API and is built into the development environment Cursor. Key change was the improved self-verification system, which reduces the number of logical errors when generating code, the model received an extended context window, allowing it to process more voluminous projects without losing coherence, developers also introduced updated security protocols that limit the generation of malicious content. The cost of using Grok 4.7 via API is $2 for 1 million input tokens and $6 for 1 million output tokens, the model demonstrates high performance, scoring 363 points in tests over 4 hours of work, which is confirmed by the data from the Hacker News community. The update is aimed at competing with current solutions for automating code writing, integration with Cursor allows developers to use the model's capabilities directly in the process of writing software, providing more accurate support for syntax and architectural solutions.