AI Battle > Rankings > Claude Fable 5.1 vs GPT-6 Astra
Head-to-head from public datasets. The better value in each row is highlighted; direction is stated per metric — for costs lower is better, for capability and speed higher is better. Pair sourced from the Humanity's Last Exam: Output Tokens per Task ranking.
| Metric | Claude Fable 5.1 | GPT-6 Astra |
|---|---|---|
| Creator | Anthropic | OpenAI |
| Intelligence index | 49.1 | 52.5 |
| Context window (tokens) | — | — |
| Input price / 1M tokens | $10 | $10 |
| Output price / 1M tokens | $50 | $50 |
| Output speed (tok/s) | — | — |
| Time to first token (s) | — | — |
Pair sourced from the Humanity's Last Exam: Output Tokens per Task ranking; leader and runner-up there.
This static page is a snapshot rebuilt daily. The interactive comparison of Claude Fable 5.1 and GPT-6 Astra loads the same numbers from the live API.
Values come from public benchmark datasets and provider list prices; see the methodology for sources, limitations and corrections. Prices are list prices — verify on the provider's pricing page before spending money.