Grok 4.7 (Low): benchmark results

Provider: xAI. Released 2026-09-21. Access: API.

Aggregate Rating 1799 ± 12, rank #147 of 2257 rated models, from 109 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Omniscience29.62Score93.5
AI for Education Pedagogy - Maths92.86Accuracy (%)93
AI for Education Pedagogy - Science93.44Accuracy (%)92.4
Wolfram LLM Benchmarking Project67.1Correct Functionality (%)90.6
AI for Education Pedagogy - Primary93.9Accuracy (%)90.5
ObviousBench99.31Answer pass³ (%)89.9
AI for Education Pedagogy89.43Accuracy (%)88.7
AA-Omniscience Accuracy49.7Accuracy (%)88.3
AutomationBench-AA - Finance - Raw Objective Completion88.28Raw Objective Completion (%)87.1
AI for Education SEND84.4Accuracy (%)86.7
AA-Briefcase - Analytical Quality Elo1721.61Elo86.6
AA-Omniscience Hallucination Rate39.93Hallucination Rate (%)86.6

Interactive version: theaggregate.ai/model?slug=grok-4-7-low · How It Works · Data refreshed daily, snapshot 2026-11-17.