SpaceXAI Unveils New Grok 4.7 Model

SpaceXAI has launched its new artificial intelligence model, Grok 4.7, aimed at enhancing its capabilities in programming and complex cognitive tasks. The model surpasses its previous version, Grok 4.6, in its ability to independently carry out longer tasks, thanks to training on a more challenging set of tasks and reinforcement learning, with a context window of up to 500,000 tokens that allows for processing extensive information within a single task. The model focuses on improving its self-verification ability, reducing hallucination rates, and features high flexibility in managing different levels of task-oriented thinking. It is particularly used in office and professional settings, with a competitive price starting at $2 per million input tokens and $6 for output tokens. Additionally, it includes capabilities to act as an intelligent agent, use external tools, search the internet, and generate reports and presentations.

Size: 120%Text Size

Notice: This Is an AI-Generated Summary

Display The Full Article