GPT-6 Astra
The reasoning heavyweight: leads the hardest exams and the cipher-cracking party tricks, and charges like it.
Key facts
- Score
- 86 / 100, rank #2
- Price
- $10 in · $50 out per 1M tokens (OpenAI API pricing)
- Blended price
- $20 per 1M tokens (3 input : 1 output)
- Context window
- 1M
- Open weights
- No
- Released
- Sep 3, 2026
- Best for
- Hard reasoning, Research-grade analysis, When cost is not the constraint
GPT-6 Astra costs $10 per million input tokens and $50 per million output tokens on OpenAI's list price (OpenAI API pricing, read Oct 10, 2026).
Independent evals
| Benchmark | Result | Who ran it | Read |
|---|---|---|---|
| Humanity's Last Exam, Diamond (Scale) | 60.6 | Scale SEAL leaderboards | Oct 9, 2026 |
| Humanity's Sixth Sense (Scale) | 53.6 | Scale SEAL leaderboards | Oct 9, 2026 |
| Artificial Analysis Intelligence Index (max) | 53 | Artificial Analysis LLM leaderboard | Oct 9, 2026 |
| FrontierMath Erdős (Epoch) | 3% (2 of 68) | Epoch AI benchmarks hub | Oct 9, 2026 |
What practitioners say
The launch drew the largest model thread of 2026 on Hacker News, and the stories that followed were about capability rather than cost: an Enigma message unsolved since 2005, a WWI cipher, a driving benchmark.
Caveats
- Long-context pricing doubles above 272K input tokens.
- Speed is the lowest of the frontier set in Artificial Analysis' measurement.
Questions people ask
How much does GPT-6 Astra cost?
GPT-6 Astra costs $10 per million input tokens and $50 per million output tokens on OpenAI's list price (OpenAI API pricing, read Oct 10, 2026).
How good is GPT-6 Astra?
GPT-6 Astra ranks #2 of 12 on the Models at Work leaderboard with a score of 86 out of 100 (October 2026 edition). The reasoning heavyweight: leads the hardest exams and the cipher-cracking party tricks, and charges like it.
What is GPT-6 Astra best for?
GPT-6 Astra is best for hard reasoning, research-grade analysis, when cost is not the constraint.
What are GPT-6 Astra's benchmark scores?
Humanity's Last Exam, Diamond (Scale): 60.6 (Scale SEAL leaderboards); Humanity's Sixth Sense (Scale): 53.6 (Scale SEAL leaderboards); Artificial Analysis Intelligence Index (max): 53 (Artificial Analysis LLM leaderboard); FrontierMath Erdős (Epoch): 3% (2 of 68) (Epoch AI benchmarks hub).
What is GPT-6 Astra's context window?
GPT-6 Astra has a 1M token context window according to OpenAI.
Is GPT-6 Astra open weights?
No. GPT-6 Astra is a proprietary model available through OpenAI's API and partner clouds.
Compare GPT-6 Astra
- GPT-6 Astra vs Claude Opus 5.5
- GPT-6 Astra vs Gemini 4 Argon
- GPT-6 Astra vs Claude Sonnet 5.5
- GPT-6 Astra vs GPT-6.1 Sol
- GPT-6 Astra vs Claude Fable 5.1
- GPT-6 Astra vs Claude Haiku 5.5
- GPT-6 Astra vs Muse Spark 1.3
- GPT-6 Astra vs DeepSeek Flash (4.1)
- GPT-6 Astra vs GPT-6 Luna
- GPT-6 Astra vs Kimi K3
- GPT-6 Astra vs Grok 4.7
Methodology: Score is 0 to 100: 50% independent evals normalised within this table, 20% price-performance, 15% practitioner sentiment from Signals and community threads, 15% operational fit (context, speed, availability, open weights). Models within three points are a tie in practice. Vendor-published figures are labelled as such and never counted as independent.