Model
Wan
Wan on UmZobo is 4 image engines and 3 video engines. Wan 3.0 text-to-video is listed at 2–30 seconds, with image-to-video and reference-to-video beside it. Token costs below are the catalog default for one output.
Image engines
| Engine | Mode | Catalog hint | Images | Tokens |
|---|---|---|---|---|
| Wan 2.7 Image Pro Spicy | t2i | Wan stills · 1K/2K | — | 15 |
| Wan 2.7 Image Pro Edit Spicy | ref2i | Up to 9 refs · edit | 1–9 | 15 |
| Wan 2.7 Image Spicy | t2i | Wan stills · 1K/2K | — | 6 |
| Wan 2.7 Image Edit Spicy | ref2i | Up to 9 refs · edit | 1–9 | 6 |
Video engines
| Engine | Type | Catalog hint | Duration | Tokens |
|---|---|---|---|---|
| Wan 3.0 T2V | t2v | Text → video · 2–30s | 2–30s | 49.50 |
| Wan 3.0 I2V | i2v | Image → video · 2–30s | 2–30s | 49.50 |
| Wan 3.0 Ref2V | r2v | 1–10 identity refs · optional audio | 2–30s | 49.50 |
Prompt notes
Use the catalog hint on the row you pick. Text-to-image engines take a prompt only. Edit and reference engines need the image count in the table. During Compare, size and quality stay at each model’s default.
Questions
What can Wan do on UmZobo?
Wan on UmZobo is 4 image engines and 3 video engines. Wan 3.0 text-to-video is listed at 2–30 seconds, with image-to-video and reference-to-video beside it. Token costs below are the catalog default for one output.
How are Wan generations priced?
Each row shows the token cost at that engine’s default size, duration, and quality. You buy token packs. There is no subscription. Compare runs of 2–5 Image Lab models add a further 25% off after the Image Lab discount.