benchmarks.site

Public leaderboard · General knowledge & reasoning

EQ-Bench

Tests how well AI models handle emotionally sensitive conversations with simulated people who have different personalities and needs.

Leaderboard

Top 3 right now

  1. 1stclaude-opus-51364.9
  2. 2ndclaude-opus-5-51339.5
  3. 3rdmuse-wonder1324.9

What it measures

Emotional intelligence in multi-turn role-play scenarios, plus separate creative-writing tests, scored by an LLM judge.

How to read the numbers

EQ-Bench 4 Elo rating (higher means judges preferred the model more often) — higher is better.

Results

25 rows · higher is better (Elo)CSV
1claude-opus-51364.91349.5 - 1378.4
2claude-opus-5-51339.51329.4 - 1349.6
3muse-wonder1324.91312.5 - 1339.0
4kimi-k31319.21308.7 - 1333.0
5claude-fable-51312.51302.8 - 1323.6
6gpt-5.51295.41285.0 - 1305.6
7claude-opus-4-71293.21283.6 - 1303.4
8gpt-6.1-sol1272.41262.9 - 1282.7
9claude-opus-4-81267.81258.6 - 1277.0
10gpt-5.41257.71250.5 - 1265.3
11muse-spark-1.11244.41230.8 - 1256.2
12gemini-4-argon1236.31225.0 - 1248.1
13gpt-5.6-sol1236.11225.4 - 1247.7
14claude-sonnet-51221.21211.6 - 1233.5
15gpt-5.6-terra1220.11209.9 - 1229.9
16Inkling1212.11201.2 - 1222.0
17claude-opus-4-61210.01198.5 - 1219.6
18glm-5.21208.11197.1 - 1216.9
19mimo-v2.5-pro1197.71185.0 - 1212.5
20claude-sonnet-4-61196.81187.9 - 1204.9
21kimi-k2.61191.41181.4 - 1200.8
22deepseek-v4-pro1156.31144.7 - 1166.7
23gpt-5.6-luna1145.91134.2 - 1156.6
24minimax-m31140.11128.8 - 1152.1
25gemini-3.1-pro-preview1132.31118.9 - 1144.4

Note: EQ-Bench 4 (120 personas, judged by Gemini 3.1 Pro, GPT-5.5, Claude Opus 4.6); sorted by Elo from site data file. Date is data generation time.

Source: EQ-Bench (Sam Paech) · eqbench.com · results as of 2026-10-24 · retrieved 2026-11-17

Markdown · More General knowledge & reasoning