Releases

Grok 4.7

SpaceXAI
MiniRouter and Grok 4.7 by SpaceXAI, separated by a silver fold of light on black.

SpaceXAI's new coding and knowledge-work model improves on Grok 4.6 in its published evaluations, with a 500K-token context window.

Grok 4.7 is not in MiniRouter’s catalog. This article covers the publisher’s announcement. Browse available models for current API access.

About the model

  • Context: 500K tokens. Vercel's advertised context and maximum output ceiling; prompt and output share the window.
  • Input: Text + image. Text output. Image understanding does not mean image generation.
  • Generation: Successor to Grok 4.6. SpaceXAI describes a larger base model and more training on longer tasks.
  • API model ID: . The Vercel route is xai; the model namespace is spacexai.

SpaceXAI publisher pricing

Publisher starting rates below exclude MiniRouter markup, long-context and fast/priority tiers. Vercel lists a separate temporary discount ending September 27; current MiniRouter rates belong to the live catalog.

SpaceXAI publisher-listed API price phases in US dollars per million tokens.
PhaseInput / 1MOutput / 1MBlended / 1MEffectiveSource
Publisher starting price$2.00$6.00$3.00Sep 21, 2026 – open-endedSpaceXAI ↗

Publisher-reported benchmarks

Results from the publisher’s announcement, grouped by unit.

BenchmarkReported resultReported byAs ofSource
DeepSWE v1.171.0%SpaceXAISep 21, 2026Announcement ↗
CursorBench 4.046.3%SpaceXAISep 21, 2026Announcement ↗
Terminal-Bench 4.038.0%SpaceXAISep 21, 2026Announcement ↗

DeepSWE v1.1 uses high effort for Grok 4.7; the other selected publisher results use xhigh. These are not MiniRouter measurements.

Vercel lists $1.20/M input and $3.60/M output at standard short-context rates on September 21, with higher rates above 200,000 input tokens and for priority service. Its promotion ends September 27, 2026.

Artificial Analysis has high and xhigh evaluations; speed and price measurements were incomplete in the September 21 feed. Missing data is left unreported.

MiniRouter availability follows the routable catalog. Upstream availability and coding-client demonstrations do not establish MiniRouter inference or client compatibility.

Independent measurements

Artificial Analysis evaluations of the first-party API.

Grok 4.7 (xhigh) ranks #7 of 462 model families on their Intelligence Index. Their headline uses the highest-scoring effort setting; every listed setting is below.

Intelligence Index
46.4
Composite of their evaluations
SettingIntelligence Index
Grok 4.7 (xhigh) ↗46.4
Grok 4.7 (high) ↗46.3

Source: Artificial Analysis ↗, read Sep 21, 2026. Measured on the model's first-party API, not a MiniRouter route.

Keep exploring

Compare published results in the model rankings and coding benchmarks, or read more model releases.

For MiniRouter product updates, visit the changelog.

Sources

  1. SpaceXAI Grok 4.7 announcement

    Published Sep 21, 2026

    Read at SpaceXAI ↗
  2. Artificial Analysis model page

    Published Sep 21, 2026

    Read at Artificial Analysis ↗

Editorial reference revised 2026-09-29. Catalog status comes from the generated routing snapshot.