GPT-6 (Low): benchmark results

Provider: OpenAI. Released 2026-09-03. Access: API.

Unified ELO 1753 ± 1, rank #26 of 1761 rated models, from 22 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Omniscience - Software Engineering (SWE)89.1Accuracy (%)98.5
AA Omniscience40.55Score98.4
AA Omniscience - Health52.6Accuracy (%)98.4
AA-Omniscience Accuracy59.53Accuracy (%)97.9
AA Humanity's Last Exam49.21Accuracy (%)97.7
AA Omniscience - Business49.8Accuracy (%)97.5
AA Omniscience - Humanities & Social Sciences56.6Accuracy (%)97.1
Artificial Analysis Intelligence Index49.32Intelligence Index96.8
AA GPQA Diamond93.13Accuracy (%)96.1
AA MMMU-Pro84.62Accuracy (%)96.1
AA Omniscience - Law57.6Accuracy (%)96
AA CritPt26.29Accuracy (%)95.8

Interactive version: theaggregate.ai/model?slug=gpt-6-low · How It Works · Data refreshed daily, snapshot 2026-09-05.