Wan 3.0 Prime Reference to Video: AI Image-to-Video Generator
Generate premium reference-guided video with Wan 3.0 Prime from up to 10 images, 5 videos, and 5 audio clips. Pay per generation.
Generate premium reference-guided video with Wan 3.0 Prime from up to 10 images, 5 videos, and 5 audio clips. Pay per generation.
About this model
Wan 3.0 Prime Reference to Video generates a higher-fidelity video guided by a prompt plus up to 10 reference images, 5 reference videos, and 5 reference audios, using the Prime tier of Wan 3.0 for sharper detail and more consistent character and style coherence. It accepts the same images_list, videos_list, and audios_list inputs as Wan 3.0 Reference to Video, mapped in array order to the reference media described in the prompt, plus 480p/720p/1080p resolution, five aspect ratios, and 2-30 second durations. For a scene generated from text alone, see Wan 3.0 Prime Text to Video; to animate a single starting image, see Wan 3.0 Prime Image to Video.
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapiapp | From $0.136 per generation (720p, 2s) up to $8.40 (1080p, 30s) | Pay-per-generation, priced by resolution and duration, no subscription required. |
| Fal.ai | Not available | Wan 3.0 Prime is not listed on Fal.ai at this time. |
| Replicate | Not available | Wan 3.0 Prime is not listed on Replicate at this time. |
Pay-per-generation, priced by resolution and duration, no subscription required.
Wan 3.0 Prime is not listed on Fal.ai at this time.
Wan 3.0 Prime is not listed on Replicate at this time.
* Competitor pricing is estimated based on similar model architectures and usage tiers.
Configuration schema
| Parameter | Type | Description | Default |
|---|---|---|---|
| Prompt | string | Describe the video you want to create. Reference media are identified by their array order. | The two people from the reference images meet on the rainy street shown in the reference video. |
| Reference Image URLs | array | Up to 10 reference images for visual coherence guidance. | undefined |
| Reference Video URLs | array | Up to 5 reference video clips (MP4/MOV, 1-15s each); total reference + generated duration must not exceed 30s. | undefined |
| Reference Audio URLs | array | Up to 5 reference audio clips (total duration up to 15s) for soundtrack synchronization. | undefined |
| Resolution | Enum (3 options) | Output video resolution. | 720p |
| Aspect Ratio | Enum (5 options) | Output video frame dimensions. | 16:9 |
| Duration | integer | Video length in seconds. | 5 |
| Thinking Mode | boolean | Enable deep-thinking mode for complex scene understanding. | false |
| Enable Audio | boolean | Include a generated audio track with the video. | true |
| Seed | integer | Random seed for reproducibility. Use -1 for a random seed. | -1 |
Describe the video you want to create. Reference media are identified by their array order.
The two people from the reference images meet on the rainy street shown in the reference video.Up to 10 reference images for visual coherence guidance.
undefinedUp to 5 reference video clips (MP4/MOV, 1-15s each); total reference + generated duration must not exceed 30s.
undefinedUp to 5 reference audio clips (total duration up to 15s) for soundtrack synchronization.
undefinedOutput video resolution.
720pOutput video frame dimensions.
16:9Video length in seconds.
5Enable deep-thinking mode for complex scene understanding.
falseInclude a generated audio track with the video.
trueRandom seed for reproducibility. Use -1 for a random seed.
-1Developer documentation
Describe the scene: Write a prompt that references your media by order, e.g. "Image1" or "Video1".
Attach reference media: Provide up to 10 URLs in images_list, 5 in videos_list, and 5 in audios_list.
Set the output shape: Choose resolution (480p, 720p, 1080p), aspect_ratio (16:9, 9:16, 1:1, 4:3, 3:4), and duration (2-30 seconds) — note total reference-video duration plus output duration must not exceed 30 seconds.
Tune generation: Enable thinking_mode for complex multi-reference scenes, and set enable_audio to false for a silent video.
Submit and poll: Send the request with your Muapi API key, store the returned request_id, then poll the prediction result endpoint or supply a webhook_url to receive the completed video URL.
curl -X POST https://api.muapi.ai/api/v1/wan3.0-prime-reference-to-video \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "The two people from the reference images meet on the rainy street shown in the reference video.",
"images_list": ["https://example.com/person1.jpg", "https://example.com/person2.jpg"],
"resolution": "720p",
"duration": 5
}'
Frequently asked
It generates a higher-fidelity video guided by a prompt plus reference images, videos, and audio for consistent characters and style.
Prime uses the same request shape and parameters but targets sharper detail and more consistent coherence across references, at a higher per-second price.
Up to 10 reference images, 5 reference videos, and 5 reference audio clips, mapped in array order to the media named in your prompt.
Yes. The total duration of any input reference video plus the generated output duration must not exceed 30 seconds.
480p, 720p, or 1080p, with five aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4).
Pricing scales with resolution and duration at the Prime rate, and the completed task response includes the exact charge for that generation.
Poll the prediction result endpoint with the returned `request_id`, or pass a `webhook_url` to be notified when the video is ready.