GPT-6 Luna (High): benchmark results

Provider: OpenAI. Access: API.

Unified ELO 1696 ± 1, rank #175 of 3363 rated models, from 63 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA-Omniscience Index - Software Engineering (SWE) - Rust68Omniscience Index91.7
AA-Omniscience Index - Software Engineering (SWE) - Go54Omniscience Index89.1
AA-Omniscience Index - Software Engineering (SWE) - Kotlin40Omniscience Index88.8
AA-Omniscience Index - Software Engineering (SWE) - Python48Omniscience Index87.7
Artificial Analysis Intelligence Index32.15Intelligence Index86.9
AA CritPt15.43Accuracy (%)86.7
AA-Omniscience Index - Software Engineering (SWE) - TypeScript45.56Omniscience Index85.6
AA Long Context Reasoning79.33Accuracy (%)84.7
AA-Omniscience Index - Software Engineering (SWE) - HTML50Omniscience Index84.6
AA-Omniscience Index - Software Engineering (SWE) - JavaScript43.64Omniscience Index84.6
AA-Omniscience Accuracy42.8Accuracy (%)83.7
AA-Omniscience Index - Software Engineering (SWE)34.2Omniscience Index83.7

Interactive version: theaggregate.ai/model?slug=gpt-6-luna-high · How It Works · Data refreshed daily, snapshot 2026-09-23.