OpenAI · model

GPT-6 Astra (medium) benchmark scores

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

As of 2026-10-07, GPT-6 Astra (medium) (OpenAI) has measured scores on 2 benchmarks and estimated scores on 2 more.

Benchmark	Score	Source
Terminal-Bench 2.1	89.5%	measured
Terminal-Bench 4.0	46.3%	estimated ± 11.6 pp, low confidence
GDP.pdf	30.4%	measured
MMMU-Pro	86.4%	estimated ± 2.6 pp, high confidence