GPT-6 Astra

GPT-6 Astra ranks #2 out of 214 models on the public BenchAlign leaderboard, with a score of 84.9/100. Its evidence status is Supported, and this profile shows 45 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.

GPT-6 Astra ranks #2 out of 173 eligible models for knowledge and understanding, with a public category score of 84.9/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.

GPT-6 Astra ranks #4 out of 144 eligible models for coding and programming, with a public category score of 76.6/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.

GPT-6 Astra has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

GPT-6 Astra ranks #1 out of 27 eligible models for reasoning and logic, with a public category score of 89.6/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.

GPT-6 Astra ranks #7 out of 120 eligible models for agentic tool use and computer tasks, with a public category score of 70.6/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.

GPT-6 Astra has source-displayable benchmark coverage for multimodal and grounded tasks, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

GPT-6 Astra has source-displayable benchmark coverage for instruction following, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

GPT-6 Astra belongs to the GPT-6 family. Related tracked variants include GPT-6.1 Sol, GPT-6 Sol, GPT-6 Luna. A sibling link indicates shared lineage or a documented configuration relationship; it does not mean the variants have identical pricing, context limits, benchmark evidence, or deployment behavior. Compare before switching.

No. GPT-6 Astra currently has 68 source-displayable rows across 655 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.

GPT-6 Astra has a documented context window of 1.05M. That figure is the maximum combined prompt and retained-conversation space reported for this exact model; it is not the maximum output length. The profile keeps output limits separate because providers often publish those limits independently.