Grok 4.7

SpaceXSI, closed weights, released Sep 2026, run at provider defaults through OpenRouter, served by SpaceXSI. It built the same five pages as every model on season 0.5 of the Rams Index, and Rams graded each one at desktop and phone size.

Rams Index, 0 to 100

66.1

95% interval 58.6–72.0

Rank: 2of 14 models by Index; the paired intervals allow 1 to 8
Cost per page: $0.13Generation only, list price on run day
Second judge: 84.5GPT-5.6 Terra places it first of 14. The Claude judge: second.
Grade sum: 25.50Seven 1–5 grades, out of 35
Pages judged: 1010 judge passes, 2 builds per brief
Defects per page: 8.0Tap targets under 24px on the phone (8 of 10)

Rank 2 of 14; the data allows anywhere from 1 to 8. Within noise of GPT-6 Astra, GPT-6.1 Sol, Claude Sonnet 5.5, GPT-6 Luna, Gemini 4 Argon, Muse Wonder and Claude Opus 5.5: the gap to each sits inside the noise, so the board can't order them yet. Not clearly behind any model. Clearly ahead of GPT-5.6 Sol, Claude Fable 5.1, Claude Fable 5, Gemini 3.8 Flash, DeepSeek V4 Pro and Claude Opus 4.8.

Designer check pending · read Nov 6Season 0.5New this seasonProvisional: 1.0 judge passes per page, protocol 2

Grades by dimension

The seven grades behind the Index, 1 to 5, mean of desktop and phone. The tick is the mean across all 14 models.

Hierarchy3.704 of 14
Typography3.751 of 14
Spacing3.301 of 14
Color3.801 of 14
Components3.555 of 14
Composure3.653 of 14
Craft3.751 of 14

Bars start at 2.8 and end at 4.2, the range the board uses, so small differences show.

Index by brief

The same Index on each of the five briefs, and where the model places on that brief among all 14.

Landing69.63 of 14
Dashboard65.22 of 14
Settings51.86 of 14
Pricing68.81 of 14
Article75.06 of 14

Bars run from 40 to 90.

How it ran

Every model runs at its provider's default settings, through the same harness, with the same budget per page.

Effort: DefaultThe provider’s default. We send no effort or thinking setting
Time per page: 284sMedian over every build, broken ones included
Output tokens: 22.1kMedian per page, reasoning included
Reasoning tokens: 13.1kMedian per page, as the API reports them
Broken pages: 0 of 10Every build came back as a whole page
Criticals: 0%0 of 10 scored pages had a critical issue in their first Rams code review
Built through: OpenRouterServed by SpaceXSI

Best and worst page

Of its 10 judged pages, the one the judge graded highest and the one it graded lowest, with the judge's top findings on each, most severe first.

Best: Article76.8
Index on this page, one judge pass. Brief mean 75.0.
- Author bio role/name order invertedSerious
- Code block horizontally overflows its containerModerate
- Blockquote mixes two headline registers without resolutionModerate
Worst: Settings51.8
Index on this page, one judge pass. Brief mean 51.8.
- Inconsistent badge styling between key rowsSerious
- Indented sub-items lack clear parent relationship signalSerious
- Inconsistent left-edge indentation between top-level and sub-items in notification listSerious

All 10 of its pages sit beside every other model's in the gallery.

Head to head

Grok 4.7 against every other model: the paired difference in Index points, and whether the 95% interval on it clears zero.

vs GPT-6 Astra-0.4Inside the noise
vs GPT-6.1 Sol+0.7Inside the noise
vs Claude Sonnet 5.5+4.3Inside the noise
vs GPT-6 Luna+4.3Inside the noise
vs Gemini 4 Argon+4.5Inside the noise
vs Muse Wonder+5.5Inside the noise
vs Claude Opus 5.5+5.7Inside the noise
vs GPT-5.6 Sol+7.1Clearly better
vs Claude Fable 5.1+7.1Clearly better
vs Claude Fable 5+8.5Clearly better
vs Gemini 3.8 Flash+10.5Clearly better
vs DeepSeek V4 Pro+11.4Clearly better
vs Claude Opus 4.8+13.2Clearly better

The other models

Season 0.5 of the Rams Index, in board order, then the unranked.

Rams Index v0.5 · updated 2026-10-17Judged by Rams eyes · Claude Sonnet 4.6