Epoch AI - Weirdml V3: leaderboard

Metric: Score. Source: epoch.ai. Saturation forecast: Around April 2028. 11 models tracked.

Top models

#ModelScore
1GPT-6 (xHigh)42.22
2Claude Opus 5.5 (xHigh)31.2
3Claude Fable 5.1 (xHigh)25.97
4Claude Opus 5 (xHigh)19.01
5Claude Fable 5 (xHigh)15.53
6GPT-5.6 Sol (xHigh)14.99
7Gemini 3.8 Flash (High)7.38
8Kimi K3 (High)7.27
9Claude Opus 4.5 (20251101) (High)6.52
10DeepSeek V4.1 Flash (High)5.95
11GPT-5.6 Luna (xHigh)5.16

Interactive version: theaggregate.ai/benchmark?slug=epoch-ai-weirdml-v3 · How It Works · Data refreshed daily, snapshot 2026-09-30.