Wan v2.5 Preview vs Sora 2
Preferred for text to video: Sora 2 · Newest: Sora 2 · Native audio: No native audio. Quality scores are blind human-preference votes; specs and prices come from each lab’s listings.
Alibaba (Qwen)
Wan v2.5 Preview
- Arena1,321
- From—
- Max length—
OpenAI
Sora 2
- Arena1,343
- From—
- Max length—
Add a model
Make it a three-way comparison.
Where they differ
Quality is the arena score from blind, side-by-side human votes. Price, length and resolution come from each lab’s own listings.
- Preferred for text to videoSora 2Wan v2.5 Preview 1,245 (#25) · Sora 2 1,343 (#17) — blind human votes
- NewestSora 2Wan v2.5 Preview Sep 24, 2025 · Sora 2 Oct 6, 2025
- Widest inputsSame inputsWan v2.5 Preview: Text prompts, Audio · Sora 2: Text prompts, Reference images
- Native audioNo native audioGenerates an audio track with the video: Wan v2.5 Preview no · Sora 2 no
- Image to videoSora 2Starts from an uploaded image: Wan v2.5 Preview no · Sora 2 yes
- Self-hostingNo open weightsBoth are available only through APIs and apps
- Most API hostsSame reachWan v2.5 Preview: 1 · Sora 2: 1
Every spec in one table
Dashes mean the data is not published.
| Specification | ||
|---|---|---|
| Quality (blind human votes) | ||
| Text to video arenaElo-style score · higher is better | 1,245 ±9#25 of 48 · 36,378 votes | 1,343 ±6#17 of 48 · 71,947 votes (best) |
| Image to video arenaElo-style score · higher is better | 1,321 ±10#22 of 48 · 16,907 votes | — |
| Output | ||
| Produces | Video | Video |
| Inputs | ||
| Text prompts | Yes | Yes |
| Start image (image to video) | No | Yes |
| Video input | No | No |
| Audio input | Yes | No |
| Access | ||
| Weights | Proprietary | Proprietary |
| API hosts | 1 | 1 |
| Released | Sep 24, 2025 | Oct 6, 2025 |
Which should you choose?
Which is better, Wan v2.5 Preview or Sora 2?
For text to video, voters prefer Sora 2 (Wan v2.5 Preview 1,245 (#25) · Sora 2 1,343 (#17)). Arena scores come from blind, side-by-side human votes on the same prompts; when the gap is inside both models’ error margins we call it a tie. Test your own prompts before you commit.
What is the difference between Wan v2.5 Preview and Sora 2?
Wan v2.5 Preview is a video model from Alibaba (Qwen) (Sep 24, 2025), taking text prompts and audio and producing video. Sora 2 is a video model from OpenAI (Oct 6, 2025), taking text prompts and reference images and producing video.
Which is newer?
Wan v2.5 Preview Sep 24, 2025; Sora 2 Oct 6, 2025.
Which generates video with sound?
Generates an audio track with the video: Wan v2.5 Preview no; Sora 2 no.
Can I run any of these on my own hardware?
No open weights. Both are available only through APIs and apps.