数据同步时间: 2026-08-17 20:11:32
排名 | 模型 | Elo 评分 Arena Elo 评分：基于用户盲测两两对决的胜负关系估算出的模型综合能力评分 | ||||
|---|---|---|---|---|---|---|
1 |  gemini-omni-flashGoogle • Proprietary | 1512±11 | ||||
2 |  flux-3-videoBlack Forest Labs • Proprietary | 1496±17 | ||||
3 |  dreamina-seedance-2.0-720pBytedance • Proprietary | 1478±9 | ||||
4 |  muse-videoMeta • Proprietary | 1457±15 | ||||
5 |  minimax-h3MiniMax • Open | 1453±14 | ||||
6 |  happyhorse-1.0Alibaba-ATH • Proprietary | 1428±13 | ||||
7 |  sora-2-proOpenAI • Proprietary | 1366±7 | ||||
8 | veo-3.1-audioGoogle • Proprietary | 1364±14 | ||||
9 | veo-3.1-audio-1080pGoogle • Proprietary | 1363±10 | ||||
10 | veo-3.1-fast-audioGoogle • Proprietary | 1362±10 | ||||
11 | veo-3.1-fast-audio-1080pGoogle • Proprietary | 1358±10 | ||||
12 | veo-3-fast-audioGoogle • Proprietary | 1347±11 | ||||
13 |  grok-imagine-video-720pSpaceXAI • Proprietary | 1347±7 | ||||
14 |  wan2.7-t2vAlibaba • Proprietary | 1343±8 | ||||
15 | sora-2OpenAI • Proprietary | 1340±7 | ||||
16 | veo-3-audioGoogle • Proprietary | 1339±13 | ||||
17 | wan2.6-t2vAlibaba • Proprietary | 1333±8 | ||||
18 | seedance-v1.5-proBytedance • Proprietary | 1256±7 | ||||
19 | veo-3Google • Proprietary | 1253±11 | ||||
20 | wan2.5-t2v-previewAlibaba • Proprietary | 1249±9 | ||||
21 | veo-3-fastGoogle • Proprietary | 1248±12 | ||||
22 | P pixverse-v5.6Proprietary • Open | 1240±11 | ||||
23 |  runway-gen-4.5Runway • Proprietary | 1223±10 | ||||
24 |  kling-2.5-turbo-1080pKlingAI • Proprietary | 1219±17 | ||||
25 | kling-2.6-proKlingAI • Proprietary | 1217±7 | ||||
显示第 1 到 25 条，共 44 条记录
每页显示:
1 / 2
排名计算方法: 系统利用在竞技场中收集到的海量双盲人类盲测对抗数据，构建 Bradley-Terry 模型以极大似然估计模型在各个单项分类下的 Elo 相对竞技值。95% 置信区间 (CI) 表明当误差区间发生重叠时，模型的真实相对实力可能不存在统计学上的显著性差异。