Best image-to-video models with audio

Still image in, video with sound out.

Minimax H3 Max (post-trained by fal) leads from Fal, ahead of 32 others.

  1. 1Fal

    Minimax H3 Max (post-trained by fal)

    1,202 elo

  2. 2ByteDance Seed

    Dreamina Seedance 2.0 720p

    1,190 elo

  3. 3MiniMax

    MiniMax H3

    1,184 elo

ranked by elo · 33 models · updated Aug 30

#ModelElo
1Minimax H3 Max (post-trained by fal)
1,202
2Dreamina Seedance 2.0 720p
1,190
3MiniMax H3
1,184
4Gemini Omni Flash
1,179
5grok-imagine-video-1.5
1,107
6HappyHorse-1.1
1,104
7MAGI-2 Preview
1,098
8Veo 3.1
1,085
9SkyReels V4
1,085
10HappyHorse-1.0
1,085
11Wan 2.7
1,082
12grok-imagine-video
1,075
13Veo 3.1 Fast
1,071
14Kling 3.0 1080p (Pro)
1,069
15Veo 3.1 Lite
1,067
16PixVerse V6
1,066
17Kling 3.0 720p (Standard)
1,065
18Vidu Q3 Pro
1,060
19Kling 3.0 Omni 1080p (Pro)
1,057
20Kling 3.0 Omni 720p (Standard)
1,049
21LTX-2.5 Fast
1,043
22Vidu Q3 Turbo
1,037
23Agnes-Video-2.5
1,033
24LTX-2.5 Pro
1,009
25Kling 2.6 Pro (January)
1,001
26Seedance 1.5 pro
1,000
27PixVerse V5.6
959
28LTX-2.3 Fast
953
29LTX-2.3 Pro
951
30LTX-2 Fast
929
31Agnes-Video-V2.0
921
32Wan 2.6
891
33LTX-2 Pro
870

what elo means

An Elo rating from head-to-head votes: two outputs are shown side by side and people pick the better one. Higher is better, and a gap of about 100 points means the higher model wins roughly two thirds of the time.