Clembench Multimodal v1.6.5: leaderboard
Three image-based dialogue games (reference game, MatchIt, MapWorld navigation) from the University of Potsdam clembench (2024); clemscore = share of games completed times mean quality score.
Metric: clemscore (imported). Source: huggingface.co. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 20 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet (20240620) | 80.77 | #298 |
| 2 | GPT-4o (2024-08-06) | 80.04 | #336 |
| 3 | GPT-4 Preview (1106) | 73.55 | #418 |
| 4 | GPT-4o (2024-05-13) | 69.56 | #325 |
| 5 | Claude 3 Opus (20240229) | 68.16 | #412 |
| 6 | Gemma 3 27B (IT) | 61.39 | #523 |
| 7 | GPT-4o Mini (2024-07-18) | 58.46 | #557 |
| 8 | Gemini 1.5 Flash | 47.73 | #617 |
| 9 | InternVL2-8B | 23.17 | #817 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=clembench-multimodal-v1-6-5 · How It Works · Data refreshed daily, snapshot 2026-10-21.