RiemannBench (Fable/Mythos): leaderboard
Mathematical reasoning benchmark reported in Anthropic's Claude Fable 5 and Claude Mythos 5 system card. This slice reports Anthropic's own runs from the Claude Fable 5 / Mythos 5 system card; small frontier-model pool.
Metric: Score (%). Source: www-cdn.anthropic.com. Saturation forecast: Around February 2028. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Mythos 5 | 55 | #2 |
| 2 | Claude Mythos Preview | 43 | #4 |
| 3 | Claude Opus 4.8 | 34 | #37 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=riemannbench-fable-mythos · How It Works · Data refreshed daily, snapshot 2026-10-20.