Standard Compute
← All models
Grok 4.7 · xAI

What does Grok 4.7 cost to run in a coding agent?

Grok 4.7 costs about $0.13 per average coding-agent request on our measured traffic, at list prices of $2 per million input and $6 per million output tokens. It scores 37.6% on Terminal-Bench 4.0, close to GLM 5.3.

List prices: xAI models and pricing, checked 2026-10-13. Request costs: Standard Compute gateway data, 79,301 coding-agent requests from 75 users, Sep 13 to Oct 12, 2026. Benchmark: Terminal-Bench 4.0 leaderboard. How we compute this

Run current Claude, GPT, Gemini and open models in your agent from $19/month

See plans

What Grok 4.7 costs per coding-agent request

Cost at list priceGrok 4.7
Prompt under 100K tokens50% of requests, 87% of input cached$0.044
Prompt of 100K to 272K39% of requests, 88% of input cached$0.16
Prompt over 272K tokens11% of requests, 86% of input cached$0.48
Average request$0.13
100 requests$13.16
Median session16 requests; a session ends after 10 idle minutes$1.21
A month of regular usemedian of 20 users active on 5+ days$222

Each logged request priced at the vendor's list rates for its own token counts, including cache reads and any higher rate for long prompts. Average output: 698 tokens per request. 1% of requests were larger than Grok 4.7's 500K context window and are left out of its figures.

What to watch for

Every token in a prompt of 200K or more costs double, and 19% of our requests reached that line. Cache reads cost $0.50 per million, a quarter of input, where most vendors charge a tenth. Its 500K window leaves out the 1% of requests that are larger.

Where Grok 4.7 sits on the Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index bar chart with Grok 4.7 (xhigh) highlighted, from artificialanalysis.ai, October 13, 2026
Artificial Analysis Intelligence Index v4.3.2, Grok 4.7 (xhigh) highlighted, downloaded October 13, 2026. Source: Artificial Analysis (artificialanalysis.ai)

About Grok 4.7

Grok 4.7 is xAI's flagship, released September 21, 2026 with a 500K-token context. It scores 46 on the Artificial Analysis Intelligence Index and 37.6% on the Terminal-Bench 4.0 leaderboard, which puts it below the current Claude and GPT-6 models and close to GLM 5.3 and GPT-5.6 Sol. Its output price ($6 per million) is low for its class, but every token in a prompt of 200K or more is billed at double rates. xAI reports 71.0% on DeepSWE v1.1 and 46.3% on CursorBench 4.0.

Sources: xAI announcement · Artificial Analysis · Terminal-Bench 4.0

Agents that run Grok 4.7 well

Output-heavy agent work at a mid-tier price.

Grok 4.7 FAQ

How much does Grok 4.7 cost per coding-agent request?
$0.13 on average at list price, $13.16 per 100 requests, and $1.21 for a median 16-request session. That is 79,301 real coding-agent requests (Sep 13 to Oct 12, 2026) priced at xAI's list rates; the median request sent 99,534 input tokens, 87% of input was read from cache, and requests returned 698 output tokens on average.
Does Grok 4.7 charge more for long prompts?
Yes. Prompts 200K tokens or more are billed at $4 per million input and $12 per million output tokens, on every token in the request including cache reads. 19% of the coding-agent requests we measured were that large.
Is Grok 4.7 good for AI agents?
Output-heavy agent work at a mid-tier price. On our editorial scoring it rates 9.0/10 for agentic ability and 9.2/10 for output quality. Low output price for its class ($6 per million).
What is Grok 4.7's context window?
Grok 4.7 has a 500K-token context window. xAI is the maker.
Can I self-host Grok 4.7?
No — Grok 4.7 is a proprietary, hosted-only model from xAI. You access it through an API (e.g. OpenRouter) rather than running the weights yourself.
What are the downsides of Grok 4.7?
The main trade-offs: Prompts of 200K tokens or more cost double; Behind the Claude 5.5 and GPT-6 models on Terminal-Bench 4.0. It's strongest for output-heavy agent work at a mid-tier price.

Grok 4.7 compared

The agent matters as much as the model. Which agent gets the most out of Grok 4.7? →

Current Claude, GPT, Gemini and open models are included with Standard Compute

Standard Compute

Current frontier models Claude Opus 5.5 · Claude Sonnet 5.5

Your agents never stop.
Your bill never grows.

Frontier models when it counts. Efficient models when it doesn’t. The most intelligence per dollar. One flat bill.

Plans from $19/mo · cancel anytime · 7-day fair refund

7-day fair refund No credit card needed Billing by Stripe1,800+ agent users