Leaked Anthropic table shows internal model scoring above Claude Mythos 5.2
Eliv_nurotic · reddit · 2026-11-10
A Reddit post shares a screenshot allegedly from an Anthropic document (section 2.3.5.3, "AECI values and trend") listing Epoch Capabilities Index scores with 95% CIs: Claude Mythos Preview 165.71, Mythos 5 168.68, Opus 5 169.40, Mythos 5.1 172.62, Opus 5.5 174.78, Mythos 5.2 177.12 — and an "Internal Model" at 178.78, topping all published models. The table notes Opus 5 and later were measured at adaptive thinking with max effort. If genuine, it suggests Anthropic already has a model stronger than Mythos 5.2 internally; authenticity is unverified.
More from Models
- GDPx leaderboard shows Fable 5.2's RSI plateauing at Sonnet 5.5 level, matching Anthropic's training caps — morgymcg · 2026-11-10
- Anthropic capped Fable 5.2's AI-research capabilities in training, system card reveals — EntropyMagnets · 2026-11-10
- Running open-weight models locally: Kimi K3.1 needs 594GB, ~$38k for 15 tok/s — ravikantagrawal · 2026-11-10
- Consultant: banks larping 'sovereignty' with on-prem OSS models; bullish on Chinese providers — teodorio · 2026-11-10
- The most complete list of LLM writing tics yet: negation-first reveals, aphoristic closers — gleech · 2026-11-10
- Amazon swaps Sonnet 4.6 for a decision model in ticket labeling, cutting cost by 98% — victormustar · 2026-11-10