Transparent pricing with no hidden fees. Pay as you go.
wan-3-0-video-primewan-3-0-video-primewan-3-0-video-prime| Model | Spec | PoYo price | Official price | You save |
|---|---|---|---|---|
| wan-3-0-video-prime | 480p · All workflows Text to Video · Image to Video · Reference to Video | $0.068/sec 13.6 credits /sec | - | - |
720p · All workflows Text to Video · Image to Video · Reference to Video | $0.140/sec 28 credits /sec | - | - | |
1080p · All workflows Text to Video · Image to Video · Reference to Video | $0.280/sec 56 credits /sec | - | - |
* Actual fees are based on the final output.
Complete guide to using Wan 3.0 Video Prime API: Text, Image, and Multimodal Reference Video
Wan 3.0 Video Prime supports text-to-video, image-to-video, and multimodal reference-to-video for workflows where generation time is a priority.
Create 2–30 second videos at 480p, 720p, or 1080p and control whether the output includes audio. Actual generation time varies by request and service conditions; no fixed generation time or speed SLA is provided.
Use three task-specific models for multimodal references, 2–30 second duration, first and last frames, and output audio control. Prime is intended for workflows where generation time matters, but it does not provide a fixed generation time or speed SLA.
01
With wan3.0-prime-reference-to-video, one request can include images, videos, and audio clips, plus either one document or one public webpage. Use the prompt to describe the required visuals, motion, and sound.
02
Set duration to any integer from 2 to 30 seconds; the default is 5 seconds. Use the supported range for product demonstrations, campaign content, and narrative scenes. Cost is calculated from the selected resolution rate and output duration.
03
Use wan3.0-prime-image-to-video with one image to define the first frame, or two images in order to define the first and last frames. Add an optional prompt to guide the action between them.
04
Use the audio field to choose whether the generated video includes an output audio track. It defaults to true and can be disabled when the workflow supplies audio separately.
Create social, advertising, and campaign video variations, then retrieve results through the asynchronous task flow.
Combine product frames, written briefs, and reference media to present features and context in one video.
Add text, frame-guided, and reference-guided video generation to customer-facing tools and SaaS products.
Submit video generation tasks asynchronously, then use task_id to retrieve status and results.
Both tiers support the same three task types and input limits. Prime is intended for workflows where generation time is a priority, while the standard tier has lower per-second rates.
| Capability | Wan 3.0 Video Prime | Wan 3.0 |
|---|---|---|
| Positioning | For workflows where generation time is a priority; no fixed generation time or speed SLA | Lower per-second rates for cost-priority workflows |
| Workflows | Text to Video · Image to Video · Reference to Video | Text to Video · Image to Video · Reference to Video |
| Output controls | 480p, 720p, or 1080p · 2–30 sec · optional audio | 480p, 720p, or 1080p · 2–30 sec · optional audio |
| Reference inputs | Up to 10 images · 5 videos · 5 audio clips · one document or one public webpage, not both | Up to 10 images · 5 videos · 5 audio clips · one document or one public webpage, not both |
| PoYo price | 480p 13.6 · 720p 28 · 1080p 56 credits/sec | 480p 10 · 720p 20 · 1080p 40 credits/sec |
| Model IDs | wan3.0-prime-text-to-video · wan3.0-prime-image-to-video · wan3.0-prime-reference-to-video | wan3.0-text-to-video · wan3.0-image-to-video · wan3.0-reference-to-video |
Both tiers use the same task types and input limits. Prime is intended for workflows where generation time is a priority, while standard Wan 3.0 has lower per-second rates. Actual generation time varies by request and service conditions.
wan3.0-prime-text-to-video, wan3.0-prime-image-to-video, or wan3.0-prime-reference-to-video. API submissions return a task_id for asynchronous status retrieval.adaptive, seed ranges from 0 to 2147483647, and enable_safety_checker defaults to true.duration to any integer from 2 to 30 seconds. The default is 5 seconds.audio to true to request an output audio track or false to disable it. The default is true.Switch between text, first/last frame, and multimodal reference generation without changing tools.
Upload images, video, audio, and one document within the published limits, or add one public webpage.
Before submission, view the estimated cost from the selected resolution rate and output duration.
Inspect the JSON request, reuse the same model ID, and use task_id to retrieve task status and generated results.