Name: 图生视频模型厂商排行榜（2026-09-15）- 最佳 AI 视频模型 - 厂商聚合排名
Creator: LM Arena
License: https://creativecommons.org/licenses/by/4.0/
Keywords: AI 图生视频模型排行榜,图生视频排行榜,最佳 AI 视频模型,AI 模型排行,Elo 积分,厂商排名,Labs

图生视频模型 Arena 榜单LM Arena Global AI Leaderboard🏆综合榜单🏆Overall

查看图生视频 AI 模型的综合 Elo 积分排名，根据输入图像生成流畅视频的效果由真实用户匿名盲测投票决定，全面反映各模型的视频生成能力。Explore Elo rankings of image-to-video AI models — smoothness and quality rated by real anonymous user votes.

2026-09-15

▾

上榜模型Models

参赛厂商Providers

3.4M

总投票数Total Votes

📖 指标说明 — 读懂这张榜单📖 Metrics Guide — Understanding the Leaderboard

🎯

榜单主指标（Score）Primary Metric (Score)

每个榜单会采用官网对应的核心排序指标。通用 Arena 榜单通常是模型竞技积分；智能体榜单则展示净提升及多项任务指标。

Each leaderboard uses the official primary ranking metric. Most Arena boards use competitive scores; the Agent board uses net improvement and task-specific metrics.

🔢

精确值（Float Value）Float Value

主指标的完整浮点数值。对于通用 Arena 榜单可用于区分整数分相近的模型；智能体榜单以百分比和置信区间展示。

The full floating-point value of the primary metric. It helps distinguish close scores on standard boards; Agent metrics are shown as percentages with confidence intervals.

📊

方差（Variance）Variance

衡量结果的统计稳定程度。绿色·稳定（<5）更可信；橙色·波动（5~12）；红色·不稳（>12）说明排名还可能明显变化。

Measures statistical stability. Green·Stable (<5); Orange·Unstable (5–12); Red·Volatile (>12) means the ranking may still move significantly.

↔️

置信区间（95% CI = ±N）95% Confidence Interval (CI = ±N)

95% 置信区间反映当前估计的不确定范围，显示为 ±N。在相同计算口径下，N 越小说明估计越集中、排名越稳定。

The 95% confidence interval, shown as ±N, summarizes uncertainty in the current estimate. Under the same methodology, a smaller N indicates a more precise estimate and a more stable ranking.

🗳️

投票数（Vote Count）Vote Count

该模型或厂商参与评测的总次数。数量越高，结果的统计置信度越高、越不容易受单次样本影响而波动。

Total evaluation volume for the model or lab. Higher counts mean higher statistical confidence and less sensitivity to individual samples.

📜

开源协议（License）License

Apache/Llama：完全开源可商用；MIT：开源限制极少；Proprietary：闭源商业模型。协议决定你能否把它集成到自己的产品里。

Apache/Llama: Fully open-source, commercial use allowed; MIT: Open-source, minimal restrictions; Proprietary: Closed-source commercial model. License determines if you can integrate it into your product.

排名Rank	名次区间Rank Range	模型Model		投票数Votes	精确分 & 区间Float Score & CI	方差Variance	价格 ($/1M)Price ($/1M)	上下文Context
🥇	1 - 3	MiniMax 最高得分模型Top Model: minimax-h3	1494±5.0	49.4K 票votes	1494.000 [1489.0, 1499.0]	2.00 · 稳定Stable	—	—
🥈	1 - 5	Google 最高得分模型Top Model: gemini-omni-1.1-flash	1488±11.0	3.7K 票votes	1488.000 [1477.0, 1499.0]	4.00 · 稳定Stable	—	—
🥉	1 - 6	阿里巴巴最高得分模型Top Model: wan3.0	1479±11.0	3.4K 票votes	1479.000 [1468.0, 1490.0]	5.00 · 波动Unstable	—	—
4	2 - 6	字节跳动最高得分模型Top Model: dreamina-seedance-2.5-720p	1475±8.0	12.2K 票votes	1475.000 [1467.0, 1483.0]	4.00 · 稳定Stable	—	—
5	6 - 9	SpaceXAI 最高得分模型Top Model: grok-imagine-video-1.5-720p	1456±4.0	130.2K 票votes	1456.000 [1452.0, 1460.0]	3.00 · 稳定Stable	—	—
6	6 - 10	HiDream 最高得分模型Top Model: hidream-o1-video-1.0	1452±12.0Preliminary	2.7K 票votes	1452.000 [1440.0, 1464.0]	4.00 · 稳定Stable	—	—
7	7 - 10	Black Forest Labs 最高得分模型Top Model: flux-3-video-20260811	1450±5.0	30.7K 票votes	1450.000 [1445.0, 1455.0]	3.00 · 稳定Stable	—	—
8	8 - 10	Alibaba-ATH 最高得分模型Top Model: happyhorse-1.0	1442±10.0	70.7K 票votes	1442.000 [1432.0, 1452.0]	2.00 · 稳定Stable	—	—
9	17 - 19	生数科技最高得分模型Top Model: vidu-q3-pro	1363±9.0	36.7K 票votes	1363.000 [1354.0, 1372.0]	2.00 · 稳定Stable	—	—
10	18 - 19	KlingAI 最高得分模型Top Model: kling-v3-pro	1355±6.0	199.2K 票votes	1355.000 [1349.0, 1361.0]	1.00 · 稳定Stable	—	—
11	32 - 40	Luma AI 最高得分模型Top Model: ray-3	1226±19.0	1.6K 票votes	1226.000 [1207.0, 1245.0]	8.00 · 波动Unstable	—	—
12	38 - 42	腾讯最高得分模型Top Model: hunyuan-video-1.5	1198±16.0	5.5K 票votes	1198.000 [1182.0, 1214.0]	4.00 · 稳定Stable	—	—
13	47 - 47	Runway 最高得分模型Top Model: runway-gen4-turbo	1052±13.0	6.8K 票votes	1052.000 [1039.0, 1065.0]	0.00 · 稳定Stable	—	—
14	48 - 48	Pika 最高得分模型Top Model: pika-v2.2	997±14.0	8.6K 票votes	997.000 [983.0, 1011.0]	0.00 · 稳定Stable	—	—

🔍

没有找到匹配的模型，换个关键词试试？No matching models found. Try a different keyword?