live
weekly refresh
basedagi.org
benchmark evidence

EQ-Bench v3

Pairwise emotional intelligence benchmark scored across eight EI dimensions. Rubric score 0-100; higher is more emotionally intelligent.

winner on EQ-Bench v3
direct benchmark result, not a broad vertical composite | source row dated 2000-01-01
scored on 2000-01-01 · stale source data (9646d)
latest mapped results | top 20
#ModelScoreEvidenceTested
1OpenAI: GPT-5.4
Openai
82.4
model-only
independent_benchmark
2000-01-01
2OpenAI: GPT-5.5
Openai
82.4
model-only
independent_benchmark
2000-01-01
3OpenAI: GPT-5.2
Openai
80.4
model-only
independent_benchmark
2000-01-01
4Anthropic: Claude Opus 4.6
Anthropic
79.2
model-only
independent_benchmark
2000-01-01
5OpenAI: GPT-5.1
Openai
78.1
model-only
independent_benchmark
2000-01-01
6DeepSeek: DeepSeek V4 Pro
Deepseek
77.6
model-only
independent_benchmark
2000-01-01
7Anthropic: Claude Opus 4.5
Anthropic
76.8
model-only
independent_benchmark
2000-01-01
8OpenAI: o3
Openai
74.7
model-only
independent_benchmark
2000-01-01
9DeepSeek: DeepSeek V4 Flash
Deepseek
72.5
model-only
independent_benchmark
2000-01-01
10Anthropic: Claude Sonnet 4.5
Anthropic
72.2
model-only
independent_benchmark
2000-01-01
11Anthropic: Claude Opus 4.7
Anthropic
68.2
model-only
independent_benchmark
2000-01-01
12Anthropic: Claude Sonnet 4.6
Anthropic
66.6
model-only
independent_benchmark
2000-01-01
13DeepSeek: R1
Deepseek
65.8
model-only
independent_benchmark
2000-01-01
14OpenAI: GPT-4.1
Openai
64.8
model-only
independent_benchmark
2000-01-01
15OpenAI: o4 Mini
Openai
59.4
model-only
independent_benchmark
2000-01-01
16Qwen: Qwen3 235B A22B
Qwen
59.1
model-only
independent_benchmark
2000-01-01
17Google: Gemma 3 27B
Google
55.6
model-only
independent_benchmark
2000-01-01
18Google: Gemini 2.0 Flash
Google
47.4
model-only
independent_benchmark
2000-01-01
19Qwen: Qwen3 32B
Qwen
46.4
model-only
independent_benchmark
2000-01-01
20Qwen: Qwen3 30B A3B
Qwen
36.7
model-only
independent_benchmark
2000-01-01
what this result means

Pairwise emotional intelligence benchmark scored across eight EI dimensions. Rubric score 0-100; higher is more emotionally intelligent.

This benchmark contributes direct public evidence. Read its scope before generalizing the result.

A win here is a win on EQ-Bench v3. Broad task pages require independent corroboration before naming a general winner.

source record
category: eq
metric: accuracy
matched models: 20
latest source date: 2000-01-01
direction: higher is better
inspect upstream source ->