Benchmark Almanac

DesignArena Image to Video

MiniMax H3 1,319
Not in model cards

Video models animate a user image into a short video. Users compare two anonymous videos and vote for the better one.

One of 40 DesignArena boards. DesignArena ranks entries with Bradley-Terry ratings from user votes.

Kind
Video generation
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Last model added
November 2026
Scores refreshed
November 12th, 2026
42
Entries on its own leaderboard, 11 models we track

MiniMax H3 leads with 1,319

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

42 entries. Scroll sideways for the rest. Bars start at 900.

  1. 1,319 MiniMax H3
  2. 1,317 Kling 4.0
  3. 1,317 P-Video-2-Pro (Quality)
  4. 1,316 HeyGen Video 1.0
  5. 1,304 Wan 3.0
  6. 1,295 FLUX 3 Max
  7. 1,281 Seedance 2.5
  8. 1,272 Grok Imagine Video 1.5
  9. 1,264 Seedance 2.0
  10. 1,256 Seedance 2.0 Fast
  11. 1,245 Happy Horse 1.1
  12. 1,242 Flux 3 Video
  13. 1,222 Grok Imagine
  14. 1,220 Seedance 2.0 Mini
  15. 1,219 HappyHorse 1.0
  16. 1,217 Gemini Omni 1.2 Flash
  17. 1,208 Gemini Omni 1.1 Flash
  18. 1,194 Gemini Omni Flash
  19. 1,194 Wan 2.7
  20. 1,185 Kling O3 (Omni)
  21. 1,181 Veo 3.1 Lite
  22. 1,178 Veo 3.1 Fast
  23. 1,170 LTX 2.5 Fast
  24. 1,170 Veo 3.1
  25. 1,156 P-Video 2
  26. 1,154 LTX 2.5 Pro
  27. 1,138 Kling v2.6 Pro
  28. 1,121 Kling v2.5 Turbo Pro
  29. 1,114 MiniMax Hailuo-2.3 (Pro)
  30. 1,111 Ray 3.14
  31. 1,101 MiniMax Hailuo-2.3 (Standard)
  32. 1,079 Sora 2
  33. 1,078 Seedance 1.0 Pro Fast
  34. 1,071 P-Video 2 Draft
  35. 1,044 Kling O1
  36. 1,030 Seedance 1.0 Lite
  37. 1,019 Seedance 1.5 Pro
  38. 1,015 Kandinsky 5.0 Pro
  39. 985 Vidu Q2 Pro
  40. 976 LTX-2 19B
  41. 951 Mischief 1
  42. 926 Cosmos 3 Super

Bars start at 900.

  1. 1 MiniMax H3 max 1,319
  2. 2 Kling 4.0 1,317
  3. 3 P-Video-2-Pro (Quality) speed 1,317
  4. 4 HeyGen Video 1.0 1080p, ultraquality 1,316
  5. 5 Wan 3.0 1,304
  6. 6 FLUX 3 Max 1,295
  7. 7 Seedance 2.5 1,281
  8. 8 Grok Imagine Video 1.5 1,272
  9. 9 Seedance 2.0 1,264
  10. 10 Seedance 2.0 Fast 1,256
  11. 11 Happy Horse 1.1 1,245
  12. 12 Flux 3 Video 1,242
  13. 13 Grok Imagine 1,222
  14. 14 Seedance 2.0 Mini 1,220
  15. 15 HappyHorse 1.0 1,219
  16. 16 Gemini Omni 1.2 Flash 1,217
  17. 17 Gemini Omni 1.1 Flash 1,208
  18. 18 Gemini Omni Flash 1,194
  19. 19 Wan 2.7 1,194
  20. 20 Kling O3 (Omni) 1,185
  21. 21 Veo 3.1 Lite 1,181
  22. 22 Veo 3.1 Fast 1,178
  23. 23 LTX 2.5 Fast 1,170
  24. 24 Veo 3.1 1,170
  25. 25 P-Video 2 1,156
  26. 26 LTX 2.5 Pro 1,154
  27. 27 Kling v2.6 Pro 1,138
  28. 28 Kling v2.5 Turbo Pro 1,121
  29. 29 MiniMax Hailuo-2.3 (Pro) 1,114
  30. 30 Ray 3.14 1,111
  31. 31 MiniMax Hailuo-2.3 (Standard) 1,101
  32. 32 Sora 2 1,079
  33. 33 Seedance 1.0 Pro Fast 1,078
  34. 34 P-Video 2 Draft 1,071
  35. 35 Kling O1 1,044
  36. 36 Seedance 1.0 Lite 1,030
  37. 37 Seedance 1.5 Pro 1,019
  38. 38 Kandinsky 5.0 Pro 1,015
  39. 39 Vidu Q2 Pro 985
  40. 40 LTX-2 19B 976
  41. 41 Mischief 1 951
  42. 42 Cosmos 3 Super 926

Source: Official DesignArena Image to Video leaderboard

#EntryScoreSettingDateSource
1 MiniMax H3 1,319 max Official leaderboard
2 Kling 4.0 1,317 Official leaderboard
3 P-Video-2-Pro (Quality) 1,317 speed Official leaderboard
4 HeyGen Video 1.0 1,316 1080p, ultraquality Official leaderboard
5 Wan 3.0 1,304 Official leaderboard
6 FLUX 3 Max 1,295 Official leaderboard
7 Seedance 2.5 1,281 Official leaderboard
8 Grok Imagine Video 1.5 1,272 Official leaderboard
9 Seedance 2.0 1,264 Official leaderboard
10 Seedance 2.0 Fast 1,256 Official leaderboard
11 Happy Horse 1.1 1,245 Official leaderboard
12 Flux 3 Video 1,242 Official leaderboard
13 Grok Imagine 1,222 Official leaderboard
14 Seedance 2.0 Mini 1,220 Official leaderboard
15 HappyHorse 1.0 1,219 Official leaderboard
16 Gemini Omni 1.2 Flash 1,217 Nov 2026 Official leaderboard
17 Gemini Omni 1.1 Flash 1,208 Official leaderboard
18 Gemini Omni Flash 1,194 Official leaderboard
19 Wan 2.7 1,194 Official leaderboard
20 Kling O3 (Omni) 1,185 Official leaderboard
21 Veo 3.1 Lite 1,181 Official leaderboard
22 Veo 3.1 Fast 1,178 Official leaderboard
23 LTX 2.5 Fast 1,170 Official leaderboard
24 Veo 3.1 1,170 Official leaderboard
25 P-Video 2 1,156 Official leaderboard
26 LTX 2.5 Pro 1,154 Official leaderboard
27 Kling v2.6 Pro 1,138 Official leaderboard
28 Kling v2.5 Turbo Pro 1,121 Official leaderboard
29 MiniMax Hailuo-2.3 (Pro) 1,114 Official leaderboard
30 Ray 3.14 1,111 Official leaderboard
31 MiniMax Hailuo-2.3 (Standard) 1,101 Official leaderboard
32 Sora 2 1,079 Official leaderboard
33 Seedance 1.0 Pro Fast 1,078 Official leaderboard
34 P-Video 2 Draft 1,071 Official leaderboard
35 Kling O1 1,044 Official leaderboard
36 Seedance 1.0 Lite 1,030 Official leaderboard
37 Seedance 1.5 Pro 1,019 Official leaderboard
38 Kandinsky 5.0 Pro 1,015 Official leaderboard
39 Vidu Q2 Pro 985 Official leaderboard
40 LTX-2 19B 976 Official leaderboard
41 Mischief 1 951 Official leaderboard
42 Cosmos 3 Super 926 Official leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard
1,217 2027 1,217 2027

Source: Official DesignArena Image to Video leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Audio and video

All 233 in Audio and video
  • Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy. Qwen 3.8 Max 88.7%
  • LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization. Gemini 4 Argon 91.7%
  • MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy. Qwen 3.8 Max 82.4%
  • MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy. Qwen 3.8 Max 90.8%
  • FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better. Phi-4-multimodal 4.00%
  • VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules. Qwen3.5-Omni-Plus 93.1%