Term
Grok 4.6
Grok 4.6 (August 2026) is xAIs flagship model for long-running agentic work — 500k-token context, text and image input, tiered pricing from 2/6 USD per million tokens.
Grok 4.6 — explained in more detail
xAI released Grok 4.6 on August 12, 2026 as its new flagship model. It builds on Grok 4.5 and focuses on long-running agents as well as more ambitious interactive and visual tasks. The model handles text and image inputs with text-only output, offers a 500k-token context window and sets no fixed limit on output length.
Example / Practical use
On the Artificial Analysis Intelligence Index, Grok 4.6 scores 61 — matching OpenAIs GPT-5.6 Sol and trailing Anthropics Claude Fable 5 by one point. Pricing is tiered: below 200k prompt tokens it is 2 USD (input) / 0.50 USD (cached input) / 6 USD (output) per 1M tokens, above that 4 / 1 / 12 USD. Grok 4.6 is available via the xAI API (Responses and Chat Completions) and in Cursor, Grok Build, OpenRouter, Vercel and Cloudflare; additionally in Microsoft Foundry, the Google Enterprise Agent Platform and Amazon Bedrock. For agent pipelines with long contexts, the combination of a large window and tiered pricing is the central practical advantage.
Distinction
Grok is the model range from xAI, the AI company in Elon Musks orbit, competing against the flagship models from OpenAI (GPT range), Anthropic (Claude) and Google (Gemini). Compared with its predecessor Grok 4.5, version 4.6 is not a full generational change but a targeted boost in agentic and visual capabilities.
Discover more
GPT-6 Astra Found Questions My First Security Review Missed
I used GPT-6 Astra and Fable 5.1 as independent reviewers for RLS, API and tenant-isolation checks. The useful part was the structured cross-review.
GlossaryGrok 4.5
Grok 4.5 is xAI's July 2026 frontier model for coding, agentic tasks and knowledge work — high token efficiency at $2 input and $6 output per 1M tokens.
EncyclopediaLLM Model Families at a Glance
Claude, GPT, Gemini, Llama & co. — who builds what, where each family shines, and how to pick the right model for your own use case.