Benchmarks / 3D
RP-Bench Community Elo
A roleplay leaderboard reporting Bayesian community Elo alongside separate multi-turn judge and behavioural evaluations. Maintained by LeviTheWeasel / lazyweasel. Score: Bayesian community Elo.
RP-Bench Community Elo
Higher is better · best effort per model
full table below
| # | Model | Effort | Score |
|---|---|---|---|
| 1 | Gemma 4 26B | xhigh | 1534 |
| 2 | Mistral Small | none | 1534 |
| 3 | Gemini 2 | – | 1529 |
| 4 | Grok 4 | high | 1517 |
| 5 | MiniMax M2 | – | 1514 |
| 6 | Sonnet 4 | high | 1513 |
| 7 | Qwen 3 | high | 1493 |
| 8 | DeepSeek V3 | none | 1492 |
| 9 | GLM 4 | – | 1490 |
| 10 | Llama 4 Maverick | medium | 1483 |
| 11 | GPT-4 | high | 1472 |