GPT-6.1 Sol (High): benchmark results

Provider: OpenAI. Released 2026-09-29. Access: API.

Unified ELO 1852 ± 17, rank #9 of 2066 rated models, from 74 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA GDP.pdf - Construction - Criterion Pass Rate84.94Criterion Pass Rate (%)100
AA GDP.pdf - Criterion Pass Rate82.86Macro Criterion Pass Rate (%)100
AI Coding Daily (Codex CLI) - Laravel Code Quality19.1Laravel Code Quality (max 20) points, LLM-judged rubric scor100
AI Coding Daily (Codex CLI) - React-TS Code Quality19React-TS Code Quality (max 20) points, LLM-judged rubric sco100
AI Coding Daily (Codex CLI) - Total58.1Total points (max 60)100
AA-Omniscience Index - Software Engineering (SWE) - TypeScript95.56Omniscience Index99.6
AA GDP.pdf32All Criteria Pass Rate (%)99.5
AA GDP.pdf - Real Estate - Criterion Pass Rate81.21Criterion Pass Rate (%)99.5
AA-Omniscience Index - Software Engineering (SWE) - Java72Omniscience Index99.4
AA-Omniscience Index - Software Engineering (SWE) - JavaScript87.27Omniscience Index99.4
AutomationBench-AA - Support - Raw Objective Completion92Raw Objective Completion (%)99
AA-Omniscience Index - Business33.1Omniscience Index98.9

Interactive version: theaggregate.ai/model?slug=gpt-6-1-sol-high · How It Works · Data refreshed daily, snapshot 2026-10-01.