xAI’s Grok 4.6 has moved into the top tier of model benchmark tables, according to The Decoder, which cites the Artificial Analysis Intelligence Index. The report says Grok 4.6 scored 61 points, tying OpenAI’s GPT-5.6 Sol and trailing only Anthropic’s Claude Opus 5 at 63 and Claude Fable 5 at 62. The result would mark a five-point gain over Grok 4.5, according to the report. That matters because the score places Grok 4.6 alongside the highest-ranked general-purpose models in the Artificial Analysis index, rather than in the next tier down. The Decoder also points to stronger performance on agentic tasks, where models carry out multi-step workflows rather than answering a single prompt. On GDPval-AA v2, a benchmark described as measuring real-world knowledge work on a computer, Grok 4.6 ranked second with an Elo score of 1,753, behind Claude Opus 5. The efficiency claim is the sharpest part of the report. The Decoder says Grok 4.6 completed complex tasks in about 53 steps, while Claude Opus 5 needed roughly 103. That is a benchmark result, not a guarantee of production behavior, but it suggests the model may be competitive in workflows where tool use and step count affect cost and latency. Pricing is the other lever. The report says Grok 4.6 remains priced at $2 per million input tokens and $6 per million output tokens. By comparison, The Decoder lists Claude Opus 5 at $5/$25 and GPT-5.6 Sol at $5/$30 per million tokens, making Grok 4.6 more than 60% cheaper on the cited pricing basis. Availability is already broad, according to the report. Grok 4.6 is available through the API, Cursor, Grok Build, and partners including OpenRouter, Vercel, and Cloudflare. The Decoder also says x.ai is offering double usage quota for the first week in Grok Build and Cursor. One caveat: this cluster contains only one report, and the source text inconsistently refers to the company as “SpaceXAI” in places and xAI/x.ai elsewhere. The benchmark and pricing details are specific, but they remain single-source within the provided material. Who benefits: Developers already using Cursor, Grok Build, OpenRouter, Vercel, or Cloudflare may get the fastest path to testing Grok 4.6. xAI benefits if benchmark parity plus lower listed pricing translates into workload migration. Who's exposed: Higher-priced frontier models are the obvious comparison point in this report, especially Claude Opus 5 and GPT-5.6 Sol. The exposure depends on whether customers see similar results outside the cited benchmarks.