Vendor · xAI
xAI LLMs
xAI builds the Grok models and sells them through the xAI API and Grok apps. Recent Grok releases emphasize coding, agentic tasks and very large context windows.
We currently review 2 xAI models. 1 of them are legacy models that have been superseded, and each is labelled with its replacement.
Family
Grok
2 versions tracked. Newest first.
Grok 4.7 is xAI's latest flagship model for coding, agentic tasks and knowledge work. It has a 500,000-token context window, accepts text and image input, and supports adjustable reasoning levels.
Legacy model, superseded by Grok 4.7. xAI's flagship model, integrated with X (Twitter) for real-time data access.
Compare versions by family
Frequently asked questions
Which xAI model should I start with?
Grok 4.7 is a sensible place to begin because it is the newest current entry we track. Compare it with the cheaper tier if you mostly need drafting and summarizing, and with the premium tier only for the hardest work.
How does xAI pricing work?
Most xAI models are sold by usage, priced per million tokens for input and output, while consumer apps use monthly plans. Each model page lists the starting price we verified.
Where can I see how xAI models compare with other vendors?
Every model has its own review, and the main LLM index at /llm/ lists all vendors side by side. Popular head-to-head pages are linked from each review.