Animate images with Wan 2.6 Spicy. Bold, high-contrast motion, optional audio-guided generation, single or multi-shot framing. Try free — pay per generation.
Alibaba WAN 2.6 Spicy Image-to-Video animates a still image into a bold, high-motion clip with optional audio-guided generation, single- or multi-shot framing, and prompt-expansion.
About this model
Wan 2.6 Spicy Image to Video turns a single still image into a bold, high-motion video clip with dramatic camera movement and vivid, high-contrast color grading. It supports optional audio-guided generation, single-shot or multi-shot framing, automatic prompt expansion for richer motion detail, and output at 720p or 1080p in 5, 10, or 15-second clips. The Spicy variant is tuned for punchier, more dynamic output than the standard Wan 2.6 image-to-video endpoint while keeping the same request shape, so switching between the two is a one-line change.
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapiapp | $1.00 – $4.50 per generation | Priced by resolution and duration: $1.00 (5s/720p) up to $4.50 (15s/1080p). Pay per generation, no subscription. |
| Fal.ai | Not available | No Spicy-tier Wan 2.6 image-to-video endpoint listed. |
| Replicate | Not available | No Spicy-tier Wan 2.6 image-to-video endpoint listed. |
Priced by resolution and duration: $1.00 (5s/720p) up to $4.50 (15s/1080p). Pay per generation, no subscription.
No Spicy-tier Wan 2.6 image-to-video endpoint listed.
No Spicy-tier Wan 2.6 image-to-video endpoint listed.
** Competitor pricing is estimated based on similar model architectures and usage tiers.
Configuration schema
| Parameter | Type | Description | Default |
|---|---|---|---|
| Image URL | string | Source image to animate into video | undefined |
| Prompt | string | Positive text prompt describing the desired motion and scene | undefined |
| Negative Prompt | string | What not to generate | undefined |
| Audio URL | string | Optional audio file to guide generation | undefined |
| Resolution | Enum (2 options) | Output video resolution | 720p |
| Duration (seconds) | Enum (3 options) | Video duration in seconds | 5 |
| Shot Type | Enum (2 options) | Single continuous shot or multi-shot sequence | single |
| Enable Prompt Expansion | boolean | Automatically expand and optimize the prompt | false |
| Seed | int | Random seed for reproducibility (-1 for random) | -1 |
Source image to animate into video
undefinedPositive text prompt describing the desired motion and scene
undefinedWhat not to generate
undefinedOptional audio file to guide generation
undefinedOutput video resolution
720pVideo duration in seconds
5Single continuous shot or multi-shot sequence
singleAutomatically expand and optimize the prompt
falseRandom seed for reproducibility (-1 for random)
-1Developer documentation
Upload a source image: provide a high-resolution still via image_url — this is the only required field.
Add a prompt (optional): describe the motion, camera movement, or scene you want; use negative_prompt to exclude unwanted elements.
Tune the output: choose resolution (720p or 1080p), duration (5, 10, or 15 seconds), and shot_type (single for one continuous shot, multi for a multi-shot sequence).
Optional extras: pass an audio_url to guide the generation with sound, set enable_prompt_expansion to let the model automatically enrich your prompt, or fix a seed for reproducible results.
Submit and poll: call the endpoint below, then poll GET /api/v1/predictions/{request_id}/result (or use a webhook_url) until status is completed.
curl -X POST https://api.muapi.ai/api/v1/wan2.6-image-to-video-spicy \
-H "Authorization: Bearer $MUAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"image_url": "https://example.com/photo.jpg",
"prompt": "dramatic slow zoom with wind-blown motion",
"resolution": "1080p",
"duration": 5,
"shot_type": "single"
}'
Frequently asked
Spicy is tuned for bolder, higher-contrast, more dramatic motion than the standard endpoint. The request and response shapes are similar, but Spicy adds `shot_type`, `enable_prompt_expansion`, and `seed` controls not present on the standard endpoint.
No — only `image_url` is required. `prompt` and `negative_prompt` are optional and let you steer the motion and scene more precisely.
720p and 1080p resolutions, with 5, 10, or 15-second clip durations. Pricing scales with both resolution and duration.
Yes — pass an `audio_url` and the model will use it to help drive the pacing and motion of the generated video.
`shot_type` controls framing: `single` produces one continuous shot, while `multi` generates a multi-shot sequence from the same source image.
Set a fixed integer `seed`. Leaving it at the default `-1` produces a different random result on every call.