Pyyan / Language Models / GPT-6 Astra
GPT-6 Astra
Saturates three of the benchmarks it reports, and ships disabled for enterprise.
currentOpenAIVerified 2026-09-05
100%
ExploitBench
Specifications
| Context | 1.05Msource · as of 2026-09-05 |
|---|---|
| Input ($/Mtok) | $10source · as of 2026-09-05 |
| Output ($/Mtok) | $50source · as of 2026-09-05 |
| Licence | Proprietarysource · as of 2026-09-05 |
| Max output | 128Ksource · as of 2026-09-05 |
| Parameters | Not publishedsource · as of 2026-09-05 |
| Released | 3 Sep 2026source · as of 2026-09-05 |
| Model ID | gpt-6-astrasource · as of 2026-09-05 |
Official links
GPT-6 Astra model pageWhat matters
- 97.6% on FrontierMath Tier 4, 99.9% on ARC-AGI-3 and 100% on ExploitBench
- Above 272K input tokens the whole request reprices: input and cache at 2x, output at 1.5x
- Rolls out to OpenAI's Daybreak cyber programme first, and arrives switched off for enterprise
Best for hard reasoning and computer use.