Grok 4.6, the xAI model in numbers

Overview

xAI released Grok 4.6 on August 12, 2026. It has a reasoning mode: it can "think" before answering, which helps with multi-step problems but uses more tokens, and therefore more money. It is a closed model: you access it through an API, from xAI or one of the 34 providers listed on models.dev. That makes it 2 months old as of October 22, 2026.

We file it under frontier models (24 models in this category on the site).

Measured performance

According to results published by Epoch AI (checked on October 22, 2026), Grok 4.6 has a capability index of 86.1 out of 100, computed on 5 benchmarks from our basket.

That puts it at #16 in our weekly model ranking.

• GPQA Diamond (PhD-level science questions): 94% — the best tracked model, Gemini 4 Argon, reaches 96.4%
• OTIS Mock AIME (competition mathematics problems): 99.2% — the best tracked model, GPT-5.5 Pro, reaches 100%
• FrontierMath (very hard research-level mathematics): 66% — the best tracked model, GPT-6 Astra, reaches 93.7%
• FrontierCode (advanced programming tasks): 48% — the best tracked model, Claude Opus 5.5, reaches 54.6%

Pricing and budget

Grok 4.6 costs $2 per million input tokens and $6 per million output tokens (developer's price, checked on October 22, 2026).

To put it in perspective: reading 20,000 tokens (about thirty pages) and writing 2,000 costs roughly $0.052; a thousand of those tasks, $52.

That is 2.4 times cheaper than the median output price of frontier models ($14.5).

Tokens read back from cache cost $0.50 per million, which helps an agent that keeps resending the same context.

To estimate your own budget, use the agent cost calculator.

Technical specifications

Its context window reaches 500,000 tokens, roughly 750 pages of text at 500 words per page: enough for a long report or several contracts at once. A single answer can run to 500,000 tokens.

It accepts as input: text, image and PDF. Its knowledge stops at February 1, 2026: beyond that, it needs web search or your documents.

It can call tools (functions, APIs) and return structured JSON output, which makes it a possible foundation for an AI agent.

When to choose it

Given its category, Grok 4.6 is mainly aimed at hard tasks where quality matters more than cost: deep reasoning, complex code, long analyses.

Its FrontierCode result (48%, above the median of tracked models) makes it a serious candidate for coding agents.

Before adopting it, test it on your own examples: benchmarks measure standard tasks, not your job.

Comparable models

In the same category, the models closest to Grok 4.6 are:

• Gemini 3.7 Flash (Google DeepMind, index 85.9)
• GPT-5.6 Terra (OpenAI, index 85.7)
• Qwen3.7 Max (Alibaba Cloud (Qwen), index 86.6)
• Gemini 3 Pro Preview (Google DeepMind, index 87)