METR: Anthropic's new internal Model 4 tops all benchmarks, may trigger 3x scaling threshold
scottleibrand · x · 2026-11-06
Citing METR's analysis, Andrew Curran reports Anthropic has a new internal model (dubbed Model 4 by METR) that finished training last month. It is the highest-scoring model in METR's set and the first to clear 50% on every benchmark tested, with scores looking linear in log-odds across 15 benchmarks. METR also guesses Anthropic has likely hit the 2x Responsible Scaling Policy acceleration threshold and may be running as high as 3x, estimating 80% of current (October 2026) acceleration is due to AI itself.
More from Models
- OpenAI launches GPT-6 Cyber, a security research model already used by Mozilla and Cisco — JiaweiLiu_ · 2026-11-06
- Pre-training done in 68 hours with 4-bit used all the way from pre- to post-training — teortaxesTex · 2026-11-06
- METR: No clear bottlenecks seen, capability progress could accelerate further in 12 months — paulnovosad · 2026-11-06
- 18 months from o3 to Claude Mythos: public models still close to last year's frontier — louisvarge · 2026-11-06
- Tencent Hunyuan Hy4 takes #1 Chinese model spot on Agent Arena at $0.57 per task — arena · 2026-11-06
- Agent Arena leaderboard: Tencent's Hy4 leads Chinese models at $0.57/task — arena · 2026-11-06