Video generation · Oct 14, 2026
Who is SOTA in Video generation?
Sora 2
Text-to-video preference from LMArena. As of Oct 14, 2026, Sora 2 is state of the art in Video generation (since Oct 17, 2025; sources LMArena Text-to-Video; methodology v1.0).
| # | Model | State | Confidence |
|---|---|---|---|
| 1 | OSora 2 OpenAI | SOTA · Video generation | medium |
| 2 | GVeo 3 Google DeepMind | Near-SOTA · Video generation | medium |
| 3 | KAKandinsky 5.0 T2v Pro kandinsky | Near-SOTA · Video generation | medium |
| 4 | GVeo 2 Google DeepMind | #4 in Video generation | medium |
| 5 | QWan V2.2 A14b Alibaba / Qwen | #5 in Video generation | medium |
| 6 | KAKandinsky 5.0 T2v Lite kandinsky | #6 in Video generation | medium |
| 7 | OSora OpenAI | #7 in Video generation | medium |
| 8 | GEMochi V1 genmo | #8 in Video generation | medium |
Embed
Badges for the leaders
[](https://isitsota.com/m/sora-2)
<a href="https://isitsota.com/m/sora-2"><img alt="SOTA · Video generation · Oct 2025" src="https://isitsota.com/badge/sora-2/video-gen.svg" height="28"></a>
Related
Best page and recent changes
Evidence recipe
What can move this crown
- LMArena Text-to-Video — primary, weight 1
Who is SOTA in Video generation?
As of Oct 14, 2026, Sora 2 is state of the art in Video generation (since Oct 17, 2025; sources LMArena Text-to-Video; methodology v1.0).
Since when is the Video generation crown current?
The current crown in Video generation has held since Oct 17, 2025.