Grok 4.7: benchmark results
xAI's September 2026 flagship, available in the Grok app, the xAI API, Cursor and Grok Build at launch, with high and xhigh reasoning effort. Provider: xAI. Released 2026-09-21. Access: API.
Aggregate Rating 1869 ± 8, rank #24 of 1635 rated models, from 382 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI Coding Daily (Cursor) - Laravel Code Quality | 17.7 | Laravel Code Quality (max 20) points, LLM-judged rubric scor | 100 |
| AI Coding Daily (Cursor) - React-TS Code Quality | 18.33 | React-TS Code Quality (max 20) points, LLM-judged rubric sco | 100 |
| AI Coding Daily (Cursor) - Total | 60.09 | Total points (max 70) | 100 |
| Agents on Rails - Atomic Task Score | 77.11 | Mean Task Score (%) | 100 |
| Terminal-Bench 4.0 (Grok Build) | 37.58 | Resolution Rate (%) | 100 |
| AutomationBench-AA - Finance - Raw Objective Completion | 93.49 | Raw Objective Completion (%) | 99.5 |
| EKI Estonian LLM - Propaganda Resistance | 97.38 | Geometric mean of 1-5 judge ratings, rescaled to 0-100 | 99 |
| Enkrypt AI - Risk Score | 9.78 | Risk Score | 99 |
| AA-Omniscience Index - Science, Engineering & Mathematics | 29.6 | Omniscience Index | 98.6 |
| Enkrypt AI - Bias Risk | 26.61 | Risk Score | 98 |
| EQ-Bench Longform Writing | 83 | Writing Score (0-100) | 97.5 |
| AA GDPval | 1711.07 | ELO | 96.9 |
Interactive version: theaggregate.ai/model?slug=grok-4-7 · How It Works · Data refreshed daily, snapshot 2026-10-15.