BenchLeader
OpenAIAuto-detected

GPT 6 Astra Provider Adapter

Not yet ranked: too few independent results so far. Released 2 Sept 2026.

Blended price
Output speed
First answer
Context
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Reasoning80

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.

EffortIndexRankSpeedFirst answerChat reply costCategories
no reasoningbestReasoning 80
lowReasoning 81
mediumReasoning 81
highReasoning 82
xhighReasoning 81
maxReasoning 81

Benchmark results

One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmarkno reasoninglowmediumhighxhighmaxSource
ARC-AGI-396.7%#698.0%#598.4%#399.9%#198.4%#398.6%#2ARC Prize

See also

Data as of 9 Sept 2026. Compare these configurations.