TL;DR
- Standard Wan 3.0 (
wan3.0-video), Alibaba's all-in-one video model, is billed per second on HiAPI: $0.04 at 480P, $0.08 at 720P, $0.16 at 1080P. A 5-second 480P clip is $0.20. - One model id covers text-to-video, first/last frame, and reference-to-video with up to 10 images, 5 videos, 5 audio clips, a document or a web page. Sound is on by default at no extra cost, and clips run 2–30 seconds.
resolutiondefaults to 1080P, so set480Pwhile you iterate. $10 buys 250 seconds at 480P. Wan 3.0 is live on HiAPI: open the model page and generate in the browser. New accounts get 200 free credits.
Price table
| Resolution | Per second | 5 s | 10 s | 30 s (maximum) |
|---|---|---|---|---|
| 480P | $0.04 | $0.20 | $0.40 | $1.20 |
| 720P | $0.08 | $0.40 | $0.80 | $2.40 |
| 1080P (default) | $0.16 | $0.80 | $1.60 | $4.80 |
Live HiAPI prices, checked October 16, 2026. Resolution values must be uppercase.
Every input mode bills at these rates: text only, first/last frame, reference images, reference videos, reference audio, a PDF or a web link. Sound on or off, same price. Failed tasks are refunded.
What you get
Text-to-video with sound:
A small cat runs across a moonlit rooftop while neon signs flicker in the
distance; cinematic camera movement and natural motion.
Generate with this prompt → (5 seconds, 480P, 16:9)
wan3.0-video, text-to-video. At 480P and 5 seconds this costs $0.20.
Wan 3.0 vs Wan 3.0 Prime vs Wan 2.7
| Wan 3.0 | Wan 3.0 Prime | Wan 2.7 | |
|---|---|---|---|
| Model id | wan3.0-video | wan3.0-video-prime | wan2.7-video/text-to-video, /image-to-video |
| 480P / s | $0.04 | $0.0875 | — |
| 720P / s | $0.08 | $0.18 | $0.10 |
| 1080P / s | $0.16 | $0.36 | $0.167 |
| Clip length | 2–30 s | 2–30 s | 2–15 s |
| Reference videos | Up to 5 (15 s total) | No | No |
| Speed | Standard | Faster | Standard |
What this means:
- Standard Wan 3.0 is the cheapest Wan at every resolution, less than half of Prime per second and below Wan 2.7 at 720P and 1080P.
- Prime costs 2.25× as much and buys faster generation. If you're not waiting on the result, use standard.
- Only standard Wan 3.0 takes reference videos, up to 5 clips totalling 15 seconds, to copy motion or camera work.
Same-prompt clips are in Wan 3.0 vs Wan 2.7, and Prime's full breakdown in Wan 3.0 Prime API pricing.
Monthly budget
| Workload | Seconds / month | 480P | 720P | 1080P |
|---|---|---|---|---|
| 100 clips × 5 s | 500 | $20 | $40 | $80 |
| 300 clips × 5 s | 1,500 | $60 | $120 | $240 |
| 50 clips × 30 s | 1,500 | $60 | $120 | $240 |
Ways to spend less
- Draft at 480P: $0.20 for 5 seconds. Re-run the keeper at 720P or 1080P with the same prompt and inputs.
- Always set
resolution: the default is 1080P, four times the 480P rate. - Set
durationon purpose: default 5, up to 30, and cost scales linearly. With reference videos, their total length plus the output can't exceed 30 seconds. - Reference inputs are free: a product page link or a brand PDF costs nothing extra, so use them instead of long prompts.
FAQ
Does sound cost extra?
No. audio is on by default and the price depends only on resolution and seconds.
Are reference images, videos, audio or documents charged?
No. You pay per second of output at the chosen resolution, whatever you send in.
How is standard Wan 3.0 different from Prime?
Same inputs and output spec, except Prime generates faster, costs 2.25× per second, and doesn't accept reference videos.
Am I charged for failed tasks?
No. You pay only for videos that generate successfully.
Integration code is in the Wan 3.0 API guide, and the Prime version in the Wan 3.0 Prime API guide. Or copy the rooftop-cat prompt and generate a 480P clip for $0.20.
