Alibaba Cloud Wan: longest single output
Alibaba Cloud Wan is read as “16 to 30 s” on the longest single output line. 8 of the sheet's 29 tools read this way; the others are BytePlus Seedance, DomoAI, LTX Studio, OpenAI Sora, SceneMixer and 2 more. From 1 source page, read 2026-09-25; no maker pays to be listed.
Without video input: an integer in the range [2, 30].
Alibaba Cloud Model Studio, Wan3.0 video generation API reference
- Wan3.0 video generation2 to 30 seconds per generation without video input; 5 seconds by defaultduration parameter
1Reading the figure
Wan3.0 accepts 2 to 30 seconds without video input, with 5 as the default and a smart mode that picks the length. With video input, input and output together may not exceed 30 seconds.
The reading goes no further than the printed figure quoted above.
| Tool | Reading |
|---|---|
| Adobe Firefly | 15 s or less |
| Alibaba Cloud Wan | 16 to 30 s |
| Amazon Nova Reel | 1 to 3 min |
| BytePlus Seedance | 16 to 30 s |
| Crreo | 10 min or more |
| DomoAI | 16 to 30 s |
| DramaSo | No figure |
| Dreamina | 1 to 3 min |
| Filmora | No figure |
| Flixly | 15 s or less |
| Google Veo | 15 s or less |
| Grok Imagine | 15 s or less |
| HeyGen | 10 min or more |
| Higgsfield | 1 to 3 min |
| Kling AI | 15 s or less |
| LTX Studio | 16 to 30 s |
| Luma | 15 s or less |
| M Studio | Uncapped, per the vendor |
| MagicLight | 10 min or more |
| MiniMax | 15 s or less |
| OpenAI Sora | 16 to 30 s |
| Pexo | 1 to 3 min |
| Pika | 15 s or less |
| PixVerse | 15 s or less |
| Revid AI | Uncapped, per the vendor |
| Runway | No figure |
| SceneMixer | 16 to 30 s |
| Topview | 16 to 30 s |
| Vidu | 16 to 30 s |
Inclusion rule. Every tool on the sheet, whether or not it has a page of its own for this line. Order. Alphabetical by tool name.