Benchmark Almanac

LMArena Image-to-Video Arena

Gemini Omni 1.2 Flash 1,503 leads since November 2026
Not in model cards

Video generation models animate the same input image with the same prompt, anonymously and side by side. Votes give Elo-style ratings.

Shares votes infrastructure with the Text-to-Video Arena. Launched on the web on January 21st, 2026.

Released
January 2026
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
47
Entries on its own leaderboard, 45 models we track

Gemini Omni 1.2 Flash leads with 1,503

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.

Bars start at 800.

  1. 1 Gemini Omni 1.2 Flash 1,503
  2. 2 Dreamina Seedance 2.5 1,494
  3. 3 MiniMax H3 1,493
  4. 4 Gemini Omni 1.1 Flash 1,487
  5. 5 Wan 3.0 1,482
  6. 6 Dreamina Seedance 2.0 1,480
  7. 7 SkyReels V4 1,475
  8. 8 Gemini Omni Flash 1,463
  9. 9 HiDream-O1-Video 1.0 1,458
  10. 10 Grok Imagine Video 1.5 1,453
  11. 11 Flux 3 Video 1,451
  12. 12 HappyHorse 1.0 1,443
  13. 13 Wan 2.7 I2V 1,427
  14. 14 Grok Imagine Video 1,411
  15. 15 Veo 3.1 Audio 1,399
  16. 16 Veo 3.1 Fast Audio 1,386
  17. 17 Vidu Q3 Pro 1,363
  18. 18 Kling v3 Pro 1,353
  19. 19 Veo 3 Audio 1,331
  20. 20 Veo 3 Fast Audio 1,327
  21. 21 Wan 2.5 I2V Preview 1,322
  22. 22 Wan 2.6 I2V 1,309
  23. 23 Seedance v1.5 Pro 1,307
  24. 24 PixVerse v5.6 1,299
  25. 25 Kling 2.6 Pro 1,294
  26. 26 Kling 2.5 Turbo 1,276
  27. 27 Seedance v1 Pro 1,274
  28. 28 Hailuo 2.3 1,263
  29. 29 Veo 3 1,259
  30. 30 Veo 3 Fast 1,259
  31. 31 Vidu Q2 Turbo 1,244
  32. 32 P-Video 1,241
  33. 33 Kling v2.1 Master 1,236
  34. 34 Hailuo 02 Pro 1,230
  35. 35 Kling v2.1 Standard 1,230
  36. 36 Ray 3 1,227
  37. 37 Hailuo 02 Standard 1,225
  38. 38 Vidu Q2 Pro 1,224
  39. 39 HunyuanVideo 1.5 1,199
  40. 40 Hailuo 02 Fast 1,195
  41. 41 Seedance v1 Lite 1,186
  42. 42 Wan 2.2 A14B 1,171
  43. 43 Veo 2 1,167
  44. 44 LTX-2 19B 1,161
  45. 45 Ray2 1,110
  46. 46 Runway Gen-4 Turbo 1,054
  47. 47 Pika v2.2 998

Source: Official LMArena Image-to-Video Arena leaderboard

#EntryScoreSettingDateSource
1 Gemini Omni 1.2 Flash 1,503 Nov 2026 Official leaderboard
2 Dreamina Seedance 2.5 1,494 Nov 2026 Official leaderboard
3 MiniMax H3 1,493 Nov 2026 Official leaderboard
4 Gemini Omni 1.1 Flash 1,487 Nov 2026 Official leaderboard
5 Wan 3.0 1,482 Nov 2026 Official leaderboard
6 Dreamina Seedance 2.0 1,480 Nov 2026 Official leaderboard
7 SkyReels V4 1,475 Nov 2026 Official leaderboard
8 Gemini Omni Flash 1,463 Nov 2026 Official leaderboard
9 HiDream-O1-Video 1.0 1,458 Nov 2026 Official leaderboard
10 Grok Imagine Video 1.5 1,453 Nov 2026 Official leaderboard
11 Flux 3 Video 1,451 Aug 2026 Official leaderboard
12 HappyHorse 1.0 1,443 Nov 2026 Official leaderboard
13 Wan 2.7 I2V 1,427 Nov 2026 Official leaderboard
14 Grok Imagine Video 1,411 Nov 2026 Official leaderboard
15 Veo 3.1 Audio 1,399 Nov 2026 Official leaderboard
16 Veo 3.1 Fast Audio 1,386 Nov 2026 Official leaderboard
17 Vidu Q3 Pro 1,363 Nov 2026 Official leaderboard
18 Kling v3 Pro 1,353 Nov 2026 Official leaderboard
19 Veo 3 Audio 1,331 Nov 2026 Official leaderboard
20 Veo 3 Fast Audio 1,327 Nov 2026 Official leaderboard
21 Wan 2.5 I2V Preview 1,322 Nov 2026 Official leaderboard
22 Wan 2.6 I2V 1,309 Nov 2026 Official leaderboard
23 Seedance v1.5 Pro 1,307 Nov 2026 Official leaderboard
24 PixVerse v5.6 1,299 Nov 2026 Official leaderboard
25 Kling 2.6 Pro 1,294 Nov 2026 Official leaderboard
26 Kling 2.5 Turbo 1,276 Nov 2026 Official leaderboard
27 Seedance v1 Pro 1,274 Nov 2026 Official leaderboard
28 Hailuo 2.3 1,263 Nov 2026 Official leaderboard
29 Veo 3 1,259 Nov 2026 Official leaderboard
30 Veo 3 Fast 1,259 Nov 2026 Official leaderboard
31 Vidu Q2 Turbo 1,244 Nov 2026 Official leaderboard
32 P-Video 1,241 Nov 2026 Official leaderboard
33 Kling v2.1 Master 1,236 Nov 2026 Official leaderboard
34 Hailuo 02 Pro 1,230 Nov 2026 Official leaderboard
35 Kling v2.1 Standard 1,230 Nov 2026 Official leaderboard
36 Ray 3 1,227 Nov 2026 Official leaderboard
37 Hailuo 02 Standard 1,225 Nov 2026 Official leaderboard
38 Vidu Q2 Pro 1,224 Nov 2026 Official leaderboard
39 HunyuanVideo 1.5 1,199 Nov 2026 Official leaderboard
40 Hailuo 02 Fast 1,195 Nov 2026 Official leaderboard
41 Seedance v1 Lite 1,186 Nov 2026 Official leaderboard
42 Wan 2.2 A14B 1,171 Nov 2026 Official leaderboard
43 Veo 2 1,167 Nov 2026 Official leaderboard
44 LTX-2 19B 1,161 Nov 2026 Official leaderboard
45 Ray2 1,110 Nov 2026 Official leaderboard
46 Runway Gen-4 Turbo 1,054 Nov 2026 Official leaderboard
47 Pika v2.2 998 Nov 2026 Official leaderboard

2 models have held number one on the official leaderboard since its release in January 2026

Lab claims count from the day the card came out, independent runs from the model’s release.

Black Forest Labs Google DeepMind
Own leaderboard 2 leaders

Source: Official LMArena Image-to-Video Arena 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
800 1,000 1,200 1,400 1,600 2027 800 1,000 1,200 1,400 1,600 2027

Source: Official LMArena Image-to-Video Arena 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 232 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%