BrowseComp (Fable/Mythos Multi-Agent): leaderboard
BrowseComp results as reported in the Claude Fable 5 / Mythos 5 system card, Multi-Agent configuration — Anthropic-run evaluation comparing a small set of frontier models.
Metric: Score (%). Source: www-cdn.anthropic.com. Saturation forecast: Around January 2027. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 | 93.3 | #16 |
| 2 | Claude Opus 4.8 | 88.5 | #37 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=browsecomp-fable-mythos-multi-agent · How It Works · Data refreshed daily, snapshot 2026-10-20.