Sora 2
LMArena ID: sora-2
OpenAIProprietary
Sora 2 is an AI video model from OpenAI. On LMArena’s Text-to-video leaderboard it ranks 17 of 48 with 1343±6 from 71,947 votes. It is not on LMArena’s Image-to-video leaderboard yet. License: Proprietary.
Key figures
- Developer
- OpenAI
- License
- Proprietary
Not yet available: Released (per AA), LMArena · Image-to-video, AA Elo
Who it is for (from the data)
- Solid quality with room to choose: #17 of 48 on Text-to-video, in the top half.
- Mainly text-to-video: no image-to-video results yet.
- Use via OpenAI's service or API: LMArena lists it as proprietary.
- Existing OpenAI users: this vendor has 3 models on Text-to-video.
Neighbours on Text-to-video
| # | Model | Score | Votes | AA Elo | License | Δ |
|---|---|---|---|---|---|---|
| 14 |  Veo 3.1 Fast Audio Google | 1362 | 39,484 | – | Proprietary | +19 |
| 15 | Veo 3.1 Fast Audio 1080p Google | 1358 | 25,870 | – | Proprietary | +15 |
| 16 | Veo 3 Fast Audio Google | 1348 | 25,258 | – | Proprietary | +5 |
| 17 | Sora 2 OpenAI | 1343 | 71,947 | – | Proprietary | |
| 18 |  Grok Imagine Video 720p xAI | 1342 | 165,462 | – | Proprietary | -1 |
| 19 | Veo 3 Audio Google | 1340 | 19,002 | – | Proprietary | -3 |
| 20 | WWan 2.7 T2V Wan | 1337 | 27,626 | – | Proprietary | -6 |
Alternatives · More from OpenAI
| # | Model | Score | Votes | AA Elo | License |
|---|---|---|---|---|---|
| 11 | Sora 2 Pro OpenAI | 1368 | 54,167 | – | Proprietary |
| 45 | Sora OpenAI | 1069 | 4,082 | – | Proprietary |
Alternatives · Top open-weight models
| # | Model | Score | Votes | AA Elo | License |
|---|---|---|---|---|---|
| 8 |  MiniMax H3 MiniMax | 1460 | 11,417 | 1120 | minimax-h3-community-license-agreement |
| 37 | KKandinsky 5.0 T2V Pro Kandinsky | 1175 | 2,019 | – | MIT |
| 38 |  Hunyuan Video 1.5 Tencent | 1169 | 4,292 | – | tencent-hunyuan-community |
| 41 | LLTX 2 19B Lightricks | 1154 | 75,596 | – | ltx-2-community-license-agreement |
| 42 |  Wan V2.2 A14B Alibaba | 1132 | 10,458 | – | Apache 2.0 |
How to read the scores
- LMArena score: people blind-pick the better of two anonymous videos for the same prompt, and the votes are fitted into an Elo-style rating; "±" is the 95% confidence interval.
- Text-to-video and image-to-video are separate boards, so scores are not comparable across them; AA Elo comes from Artificial Analysis on a different scale.
- If the gap between two models is smaller than their combined intervals, the order is not statistically reliable; treat them as the same tier. More votes mean narrower intervals.
FAQ
Where does Sora 2 rank among AI video models?
Text-to-video rank 17 (1343) (LMArena, data as of Oct 27, 2026).
Is Sora 2 open-weight?
No. LMArena lists it as Proprietary; it is available only through OpenAI’s products or API.
Can Sora 2 turn an image into video?
Who makes Sora 2?
According to LMArena, Sora 2 is made by OpenAI.
How reliable is Sora 2's score?
The scores are based on 71,947 votes on Text-to-video (95% CI ±6). More votes and narrower intervals mean a more stable rank.
How far is Sora 2 from #1 on Text-to-video?
#1 on Text-to-video is Gemini Omni 1.1 Flash (1516); Sora 2 scores 1343, 173 points behind.
Sources
LMArena Text-to-video data published on Sep 22, 2026.
LMArena Image-to-video data published on Sep 22, 2026.
17NAS syncs the upstream leaderboards daily; this page's data was synced on Oct 27, 2026.
Price, clip length, resolution and speed have no reliable public data yet and are omitted; all values come from the snapshot, blanks show –.