AI Pulse · AI Trends Pulse

xAI just released Grok 4.6, and the pricing makes this one worth paying attention to. It scores 61 on the Artificial Analysis Intelligence Index, matching OpenAI’s GPT-5.6 Sol and sitting just one point behind Anthropic’s Claude Fable 5 Max at 62. That puts it in the top tier of current models.

The interesting part is cost. Grok 4.6 runs at $2 per million input tokens and $6 per million output tokens. Sol charges $5 and $30 for the same volumes. You’re getting comparable performance at roughly a third of the price. For a business running meaningful volume through an API, that difference compounds quickly. If you’re processing customer support threads, analyzing documents, or generating content at scale, the math starts to matter.

The model is already live in Cursor, the AI code editor, and in Grok Build, xAI’s development environment. That means you can start testing it today if you’re already working in those tools. Artificial Analysis has published the full benchmarks and breakdown, and the numbers hold across reasoning, coding, and general language tasks.

This is the kind of release that changes vendor conversations. If you’ve been locked into a single provider because performance mattered more than cost, that trade off just shifted. It’s also exactly the sort of update you want surfaced automatically in a system like our Omni Command Centre, where model performance, pricing, and availability get tracked in one place so you’re not hunting through release notes every week.

The model isn’t perfect. It’s tied with Sol, not ahead of it, and it’s still a point behind Claude’s best. But for most real world tasks, that gap won’t show up in your output quality. The price gap will show up in your invoice.