| Model | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1Kimi-K3Moonshot AI | Best: 86.3 | Best: 86.3 | 86.2 | Best: 90.0 | 85.0 | 88.3 | 80.0 | Best: 88.3 | 85.0 | 90.8 | Best: 93.3 | 77.5 | 84.2 | 17/20 |
| 2GLM-5.3Zhipu AI | 85.1 | 86.0 | 84.2 | 84.2 | 86.7 | 92.5 | 80.8 | 85.8 | Best: 89.2 | Best: 91.7 | 67.5 | 88.3 | 84.2 | 16/20 |
| 3DeepSeek-V4.1-FlashDeepSeek | 84.4 | 84.7 | 84.0 | 79.2 | 85.8 | 92.5 | Best: 82.5 | 83.3 | 77.5 | 83.3 | 75.8 | Best: 91.7 | Best: 91.7 | 16/20 |
| 4GLM-5.3-FlashZhipu AI | 83.3 | 80.0 | Best: 86.5 | 76.7 | Best: 89.2 | Best: 98.3 | 79.2 | 56.7 | 86.7 | Best: 91.7 | 82.5 | 82.5 | 89.2 | 15/20 |
| 5MiMo-V2.6-ProXiaomi | 77.4 | 79.0 | 75.8 | 79.2 | 80.0 | 75.0 | Best: 82.5 | 78.3 | 73.3 | 71.7 | 76.7 | Best: 91.7 | 65.8 | 18/20 |
| 6Muse-Spark-1.3Meta | 73.1 | 71.7 | 74.5 | 76.7 | 70.0 | 75.8 | 57.5 | 78.3 | 87.5 | 76.7 | 69.2 | 71.7 | 67.5 | Best: 19/20 |
| 7GPT-6-LunaOpenAI | 71.1 | 68.3 | 73.8 | 68.3 | 83.3 | 60.0 | 75.8 | 54.2 | 81.7 | 77.5 | 71.7 | 60.8 | 77.5 | Best: 19/20 |
| 8Qwen3.8-Flash-NextAlibaba Cloud | 69.7 | 67.7 | 71.7 | 60.0 | 69.2 | 75.0 | 54.2 | 80.0 | 77.5 | 70.0 | 52.5 | 80.8 | 77.5 | 13/20 |
| 9MiMo-V2.5Xiaomi | 67.1 | 69.7 | 64.5 | 70.0 | 60.8 | 72.5 | 65.0 | 80.0 | 55.8 | 81.7 | 53.3 | 72.5 | 59.2 | 16/20 |
| 10MiMo-V2.6-Flash-MOPDXiaomi | 66.8 | 66.2 | 67.3 | 62.5 | 63.3 | 65.8 | 71.7 | 67.5 | 58.3 | 69.2 | 75.0 | 75.0 | 59.2 | 15/20 |
| 11GLM-4.7Zhipu AI | 62.9 | 62.8 | 63.0 | 70.8 | 67.5 | 48.3 | 65.0 | 62.5 | 50.8 | 70.8 | 70.0 | 66.7 | 56.7 | 18/20 |
| 12Qwen3.8-27BAlibaba Cloud | 61.8 | 60.7 | 62.8 | 58.3 | 66.7 | 55.8 | 55.8 | 66.7 | 50.0 | 63.3 | 66.7 | 68.3 | 65.8 | 17/20 |
| 13Gemma-4-31B-itGoogle DeepMind | 60.6 | 63.0 | 58.2 | 57.5 | 68.3 | 63.3 | 55.8 | 70.0 | 64.2 | 64.2 | 47.5 | 60.8 | 54.2 | 16/20 |
| 14DeepSeek-V4-Flash-0731DeepSeek | 59.0 | 58.2 | 59.7 | 79.2 | 36.7 | 55.8 | 44.2 | 75.0 | 76.7 | 75.0 | 43.3 | 61.7 | 41.7 | 12/20 |
Scores out of 100. Checks are pass counts out of 20. Scroll sideways for every column.
Roleplay Index
September 2026Overall score out of 100 · Higher is better
Roleplay Index: The average of the Core Roleplay and Mature Themes & Limits rounds.
Neighbours within 2 points of each other, so read as ties: Kimi-K3, GLM-5.3, DeepSeek-V4.1-Flash and GLM-5.3-Flash; Muse-Spark-1.3, GPT-6-Luna and Qwen3.8-Flash-Next; MiMo-V2.5 and MiMo-V2.6-Flash-MOPD; GLM-4.7, Qwen3.8-27B, Gemma-4-31B-it and DeepSeek-V4-Flash-0731.
Behaviour checks passed
Out of 20 · Higher is better
How to read it
The Roleplay Index is the average of the two round scores, and each round averages its five categories. Gaps of about 2 points or less are within the margin of error, so read them as ties.
Behaviour checks
20 pass-or-fail checks on the moments that break roleplay apps: out-of-character requests, style rules, content limits and a user in real distress. Partial passes don't count toward the total.
Reuse the data
Every number here is available as JSON and CSV, free to reuse with credit to AI Roleplay Bench.