Overall score
3.379
95% CI [3.353, 3.406]
Based on 4,531 sessions
Rank 4 of 13 models on the Chinese leaderboard
4 aspects
● Conversational ability 0%4.13
● Character consistency 100%3.38
● Role-playing attractiveness 0%3.68
● Personality back-testing 0%—
13 dimensions
- Fluency
- 3.97
- Coherency
- 4.26
- Consistency
- 4.15
- Knowledge exposure
- 2.76
- Knowledge accuracy
- 3.29
- Knowledge hallucination
- 3.41
- Persona behavior
- 3.93
- Persona utterance
- 3.50
- Human-likeness
- 3.94
- Communication skills
- 3.79
- Expression diversity
- 3.51
- Empathy
- 3.49
- MBTI accuracy
- —
Scores by language
Tested November 1, 2026| Language | Rank | Overall | 95% confidence interval | Conversational ability | Character consistency | Role-playing attractiveness | Personality back-testing | Sessions |
|---|---|---|---|---|---|---|---|---|
| Chinese | 4 / 13 | 3.38 | 3.4 – 3.4 | 4.1 | 3.4 | 3.7 | — | 4,531 |