Introduction
Wan 3.0 Pro Prime Text to video generates a 2 to 30 second shot from a prompt and delivers it at 1080p, 2K or 4K with audio in the same file. Its field list matches Wan 3.0 Pro Text to video exactly, but the base render is Wan 3.0 Prime, which Alibaba positions as the high-speed version with the same capabilities. A native Prime render tops out at 1080p, so the 2k and 4k tiers are enlarged from it rather than rendered natively.
Choose a Wan 3.0 Pro Prime mode by starting material
| What you have at the start | Mode |
|---|---|
| a written scene or script | Wan 3.0 Pro Prime Text to video |
| an approved opening image | Wan 3.0 Pro Prime Image to video |
| separate visual, motion, or sound references | Wan 3.0 Pro Prime Reference to video |
Key features & highlights
The same body as Wan 3.0 Pro
prompt (up to 20,000 characters) is the only required field; resolution defaults to 1080p, aspect_ratio to adaptive and enable_prompt_expansion to true. Because the contract is identical, one JSON body runs on both Pro and Pro Prime with only the model name changed, so send a test shot to each with the same seed before committing a project to one of them.
Framing written into the brief
The crypt example spends a third of its prompt on framing alone: the whole head in frame, clear headroom above the hair, the frame cut at the waist, the face visible at all times. In a close-up the usual failure is the model cropping the face out of its own frame, and naming the framing is the fix. Candlelight is the only key, and the audio — flames guttering, a far-off iron door — arrives in the same pass.
Motion with nothing to settle on
The underwater example has no ground to reference: a crimson silk gown blooming as the subject sinks, hair fanning out, bubbles trailing from her fingertips, and a camera that arcs and tilts at the same time. It sets no resolution, so it runs at the default 1080p. Shots in which nothing settles are the ones that benefit most from a larger delivery tier.
Request parameters
| Parameter | Required | Description |
|---|---|---|
prompt | Yes | Subject, one action, one camera move, then light and sound. At 4K a vague brief costs real money, so write the shot rather than a mood. at least 1 characters · at most 20000 characters |
resolution | No | 1080p, 2k or 4k, default 1080p. |
aspect_ratio | No | 16:9, 4:3, 1:1, 3:4, 9:16 or adaptive. Defaults to adaptive, which lets the model choose; name a ratio when the delivery format is fixed. |
duration_seconds | No | Any whole number from 2 to 30, default 5, or -1 for smart duration, where the model picks the length: the 30-second maximum is reserved up front and the charge settles on the le… |
generate_audio | No | On by default. Sound is generated with the picture and arrives inside the same file. Default: |
enable_prompt_expansion | No | On by default on this family. It fills out thin briefs and rewrites detailed ones, so turn it off once your prompt is a precise shot description. Default: |
seed | No | 0 to 2147483647. Fix it while you tune resolution or duration, otherwise every rerun changes two things at once. |
Pricing
| Specification | Price |
|---|---|
| $0.234 / second |
| $0.252 / second |
| $0.279 / second |
Use cases
Moving coloured light in a vertical frame
The aurora example asks for green light running across skin and hair as two people share one blanket on a frozen lake, framed 9:16. Moving coloured light across skin is where banding and colour noise show first, so commission 2K or 4K for a subject like this only when the scene actually needs it; a static subject at 4K only buys a larger file.
Handheld, available-light scenes
The night-ferry example is deliberately plain: handheld documentary framing, a deck floodlight and distant harbour lights, two adult men at the stern rail pressing their foreheads together and kissing as the wake churns below. There is no content flag to set, since every scene is sent like any other prompt; consent, rights, age and local law remain your responsibility.
Prompt guide & pro tips
Explore small, deliver large
Tune the prompt at 1080p and 5 seconds with seed fixed. Once the take works, re-send the identical body with resolution set to 4k. Changing resolution and seed together changes two things at once and hides which one moved the result.
Check hard light first in an enlarged file
A single hard source at night — one hanging bulb, one deck light — is where enlarged footage shows its limits first: banding in the falloff, waxy skin, wet fabric turning into a grey smear. Review those areas in the delivered 2K or 4K file before grading or cutting a sequence from it.
Lock the wording before the final render
Prompt expansion is on by default on this family and rewrites detailed briefs. Turn enable_prompt_expansion off once the prompt reads as a precise shot description, so the version you tuned is the version that runs. Write the subject, one action, one camera move, then light and sound.
API quick start
# Generate the idempotency key once; reuse it when retrying after a lost response.
export SPICY_IDEMPOTENCY_KEY="${SPICY_IDEMPOTENCY_KEY:-$(uuidgen)}"
curl -X POST https://api.spicyapi.ai/api/v1/jobs/createTask \
-H "Authorization: Bearer $SPICY_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: $SPICY_IDEMPOTENCY_KEY" \
-d '{
"model": "alibaba/wan-3.0-pro-prime/text-to-video",
"input": {
"prompt": "A slow aerial shot drifting over a misty pine forest at sunrise",
"resolution": "1080p",
"aspect_ratio": "adaptive",
"duration_seconds": 5
}
}'
# Copy data.taskId from the response into TASK_ID, then poll the status.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.spicyapi.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
-H "Authorization: Bearer $SPICY_API_KEY"
See the API reference for authentication, the complete schema, polling, and error handling.
