Tongyi Wanxiang 3.0. 2–30s, 720p / 1080p, up to 10 references.
Wan 3.0 APIActive
Resolution pricing
Billed by resolution tier at the current list price (discount included).
| Tier | Unit price | Discount | 8-second clip |
|---|---|---|---|
| 720p | ¥0.586/s | 2% off | ¥4.698 s output |
| 1080p | ¥0.987/s | 17% off | ¥7.98 s output |
With reference video, reference seconds are billed too, at the same unit price as the selected resolution (720p / 1080p); reference videos total up to 15 seconds.
Each reference video's duration is read separately, rounded up to whole seconds, then summed (e.g. 3.2 s + 0.6 s = 5 s). Requests whose reference videos exceed the cap, or that ask for an unlisted tier, are rejected (400) and not charged.
8-second sample: standard tier = unit price × 8. Actual charges follow your billing details.
Related Models
More models from 图片生成
Compare models from other providers
Wan 3.0 API: price, model ID, and how to call it
Call Wan 3.0 through AI Code With. Use model ID wan-3.0. billed per second by resolution tier: 720p ¥0.586/s, 1080p ¥0.987/s.
Next: Get API key · See integration guides · All model prices · All model prices
FAQ
What is the Wan 3.0 model ID?
Send model "wan-3.0" in the API request. The page path /wan-3.0 is only the URL, not the model parameter.
How do I call the Wan 3.0 API?
Generate it on the AI video page, or call the video generation API with the same AI Code With API key and set model to wan-3.0.
How is Wan 3.0 priced?
Wan 3.0 is billed per second by resolution tier: 720p ¥0.586/s, 1080p ¥0.987/s. An 8-second 720p clip costs about ¥4.69. With reference video, reference seconds are billed too, at the same unit price as the selected resolution (720p / 1080p); reference videos total up to 15 seconds. Each reference video's duration is read separately, rounded up to whole seconds, then summed (e.g. 3.2 s + 0.6 s = 5 s). Requests whose reference videos exceed the cap, or that ask for an unlisted tier, are rejected (400) and not charged. Actual charges follow your billing details.
GPT-5.6 Sol