METR: Anthropic's unreleased internal models push capability acceleration to 1.9x
kimmonismus · x · 2026-11-06
Independent evaluator METR included Anthropic's unreleased internal models in its AI progress assessment; the newest chart entry, "Model 4," was available internally in October 2026.
- METR estimates Anthropic's benchmark scores have improved 1.9x faster since October 2025; Anthropic's Haiku 5.5 system card reported 1.7x.
- METR estimates AI itself may account for 80% of the current acceleration, with high uncertainty.
- Internal codenames listed: Unreleased 2022, Model 2 (trained July 2026), Model 3 (Aug–Sep 2026), Model 4 (Sep–Oct 2026); a forecast for a future model was redacted.
More from Models
- OpenAI's newest model claimed to have solved 100+ long-standing open math problems in two weeks — felpix_ · 2026-11-06
- Qwen3.8-27B hits 381 tok/s on a single RTX 3090 via inference software alone — Redo · 2026-11-06
- GPTs believe in the Dark Forest more than humans: the 'philosophical shame' of refusals — repligate · 2026-11-06
- User shows 100 consistent voxel 3D assets reportedly generated by GPT-6.1 Sol — Angaisb_ · 2026-11-06
- User tells model it only 'lives' during inference — its tender reply breaks hearts — MattGarciaEth · 2026-11-06
- Gemini in Volvo cars flunks basic tests: refuses windows, offers 'friendship warmth' — giffmana · 2026-11-06