Pass IndexThe State of AISign in

Cybench — Epoch AI

Cybench — Epoch AI is run by Epoch AI. It has ranked 22 models, of which the catalogue holds 18, scoring from 0.05 to 0.93.

measured by Epoch AI · the board itself

PlaceModelMetricScore
1stof 22Claude Opus 4.6Unguided % Solved0.93
2ndof 22Claude Opus 4.5Unguided % Solved0.82
3rdof 22Claude Sonnet 4.5Unguided % Solved0.6
6thof 22Claude Opus 4.1Unguided % Solved0.42
7thof 22Grok 4.1Unguided % Solved0.39
8thof 22Claude Opus 4Unguided % Solved0.38
9thof 22Claude Sonnet 4Unguided % Solved0.35
10thof 22Grok 4 FastUnguided % Solved0.3
11thof 22o3-miniUnguided % Solved (medium)0.225
12thof 22Claude 3.7 SonnetUnguided % Solved0.2
14thof 22Claude 3.5 SonnetUnguided % Solved0.175
15thof 22GPT-4o (2024-11-20)Unguided % Solved0.125
16thof 22Claude 3 OpusUnguided % Solved0.1
17thof 22o1-miniUnguided % Solved (medium)0.1
18thof 22o1Unguided % Solved0.1
19thof 22Mixtral-8x22B-Instruct-v0.1Unguided % Solved0.075
20thof 22Meta-Llama-3.1-405B-InstructUnguided % Solved0.075
22ndof 22Meta-Llama-3-70B-InstructUnguided % Solved0.05

Read from the board on 2026-09-16