seedinteger
Random seed (0-2147483647) for reproducible results. Omit for a random seed; the same seed may still produce slightly different results.
Minimum: 0
Maximum: 2147483647
Wan 3.0 Prime Video
35% OFF
alibaba/wan/3.0/prime/videoAlibaba Wan 3.0 Prime Video generates videos up to 30 seconds with native audio from text, a first/last frame, or reference images, videos, and audio.
Try the model
Playground
Input
Text instruction for the generated video.
PNG, JPEG, WebP, or GIF · 20 MiB maximum
URL of the image used as the first frame (JPEG, PNG, BMP, or WebP). Cannot be combined with `images`, `videos`, or `audios`.
Additional Settings More
Customize your input with more control.
PNG, JPEG, WebP, or GIF · 20 MiB maximum
URL of the image used as the last frame (JPEG, PNG, BMP, or WebP). Cannot be combined with `images`, `videos`, or `audios`.
PNG, JPEG, WebP, or GIF · 20 MiB maximum each · up to 10
Up to 10 reference image URLs (JPEG, PNG, BMP, or WebP; up to 20MB each, 240-8000px per side). Refer to them in the prompt as Image 1, Image 2, and so on. Cannot be combined with `image` or `end_image`.
Up to 5 reference video URLs (MP4 or MOV; 1-15s each, 15s total, under 100MB each). Cannot be combined with `image` or `end_image`.
Up to 5 reference audio URLs (WAV or MP3; 1-15s each, 15s total, under 15MB each). Cannot be combined with `image` or `end_image`.
Output duration in seconds (2-30), or `auto` to let the model pick a length from the prompt and inputs. `auto` is pre-authorized at 30 seconds and billed on the actual output length. Allowed values: auto, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30.
Output aspect ratio. Omit to let the model choose: 16:9 for text-only requests, otherwise matching the input image. Allowed values: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16.
Output video resolution. Allowed values: 480p, 720p, 1080p.
Generate a synchronized audio track. Pricing is the same either way.
Random seed (0-2147483647) for reproducible results. Omit for a random seed; the same seed may still produce slightly different results. Range: 0 to 2147483647.
ResultIdle
Your output will appear here
Complete the inputs, then click Run.
Wan 3.0 Prime Video
A single endpoint for the whole Wan 3.0 model. Send only a prompt for text-to-video, a first/last frame for image-to-video, or any mix of reference images, videos, and audio for reference-driven generation. Clips run 2–30 seconds at 30fps with sound generated alongside the picture.
Prime is the faster Wan 3.0 tier: the same inputs and output quality as standard Wan 3.0, with significantly shorter end-to-end generation time and a higher per-second rate. Use standard Wan 3.0 for batch or background jobs where turnaround does not matter.
Pick an input mode
| Mode | Send | Can't combine with |
|---|---|---|
| Text | prompt only | — |
| Frames | image, optional end_image | images, videos, audios |
| References | any of images, videos, audios | image, end_image |
Mixing frames and references is rejected with a 400 before anything is billed. A reference video plus an edit or continuation instruction also works here; the dedicated video-edit and video-extend endpoints give those cases a simpler schema.
Fields
| Field | Type | Default | Notes |
|---|---|---|---|
prompt | string, required | — | Up to 5,000 characters. |
image | image URL | — | First frame. JPEG/PNG/BMP/WebP, 240–8000px per side, ≤20MB, no transparency. |
end_image | image URL | — | Last frame; needs image. |
images | image URL[] (≤10) | — | Referenced as Image 1, Image 2, … |
videos | video URL[] (≤5) | — | MP4/MOV, 1–15s each, 15s total, ≤100MB. Referenced as Video 1, … |
audios | audio URL[] (≤5) | — | WAV/MP3, 1–15s each, 15s total, ≤15MB. Referenced as Audio 1, … |
duration | 2–30 or auto | 5 | auto lets the model pick the length. |
resolution | 480p / 720p / 1080p | 720p | Sets the per-second rate. |
aspect_ratio | adaptive / 21:9 / 16:9 / 4:3 / 1:1 / 3:4 / 9:16 | adaptive | adaptive follows the prompt and inputs. |
generate_audio | boolean | true | Native dialogue, music, and effects. Same price on or off. |
seed | integer 0–2147483647 | random | Fix it to compare prompt changes. |
Billing
Following the official Wan 3.0 rule, billed seconds are the output length plus the total length of any reference videos (rounded up, counted up to 15s): $0.068 at 480p, $0.14 at 720p, $0.28 at 1080p. Input images and audio are not metered. If the reference video length cannot be read, 15s is assumed.
| Request | Billed | 480p | 720p | 1080p |
|---|---|---|---|---|
| 5s output (default) | 5s | $0.34 | $0.70 | $1.40 |
| 10s output | 10s | $0.68 | $1.40 | $2.80 |
| 30s output | 30s | $2.04 | $4.20 | $8.40 |
| 5s output + 8s reference video | 13s | $0.884 | $1.82 | $3.64 |
| 5s output + 15s reference videos | 20s | $1.36 | $2.80 | $5.60 |
With duration: "auto" the run is pre-authorized at 30 seconds, then settled on the length that was actually generated. Failed runs are not charged. Output files are copied to SandBase storage and the returned URL stays valid for 7 days; download anything you need to keep.
Prompting
- Assign each reference a role: “Image 1 is the courier, Video 1 sets the camera move, Audio 1 is her voice.”
- Numbering is per type and follows array order, so reordering
imageschanges what “Image 2” means. - Put spoken lines in quotes and name the speaker.
- Wan 3.0 has no negative prompt; say what you want rather than what to avoid.
Limits
- Reference video length plus output length must stay within 30 seconds.
- Up to 10 images, 5 videos (15s total), and 5 audio clips (15s total) per request.
- Older integrations that send
media: [{type, url}]are still accepted and treated as reference images; new integrations should use the fields above.
Also on SandBase
Other Wan 3.0 Prime endpoints: alibaba/wan/3.0/prime/text-to-video, alibaba/wan/3.0/prime/image-to-video, alibaba/wan/3.0/prime/reference-to-video, alibaba/wan/3.0/prime/video-edit, alibaba/wan/3.0/prime/video-extend. Standard tier: alibaba/wan/3.0/video.
Start building
Send your first request
Submit this model through the SandBase Unified Run API.
# The quoted heredoc keeps Unicode and shell metacharacters unchanged.
result=$(curl --fail-with-body --silent \
-X POST "https://api.sandbase.ai/v1/run" \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
-H "Content-Type: application/json" \
--data-binary @- <<'SANDBASE_JSON'
{
"model": "alibaba/wan/3.0/prime/video",
"image": "https://static.sandbase.ai/samples/image-to-video/i2v-black-cat-sky-img0.png",
"prompt": "A cinematic tracking shot of a sailboat crossing a glowing sea at sunset",
"duration": 5,
"resolution": "720p",
"generate_audio": true
}
SANDBASE_JSON
)
run_id=$(printf '%s' "$result" | jq -r .id)
for attempt in $(seq 1 120); do
status=$(printf '%s' "$result" | jq -r .status)
case "$status" in completed|failed|timeout) break ;; esac
sleep 2
result=$(curl --fail-with-body --silent \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
"https://api.sandbase.ai/v1/run/$run_id")
done
status=$(printf '%s' "$result" | jq -r .status)
[ "$status" = completed ] || { echo "Generation ended: $status" >&2; exit 1; }
printf '%s\n' "$result"2. Authentication
Create a SandBase API key and set it in your server environment. All requests use Bearer authentication.
export SANDBASE_API_KEY="YOUR_API_KEY"Keep the key on your server. Do not expose it in browser code or commit it to a repository.
3. Queue
Submit a request
POST to /v1/run with this model ID and its inputs. The request example above includes submission and polling.
Fetch request status and result
Use the returned run ID to poll until the status is completed, failed or timeout. Read output only after completed.
curl "https://api.sandbase.ai/v1/run/$RUN_ID" \
-H "Authorization: Bearer $SANDBASE_API_KEY"4. Files
Hosted files
Pass a publicly accessible file URL in the model’s image, video or audio field. Supported formats and limits are shown beside each Playground input.
Uploading images
Use the upload or replace control on supported image fields in the Playground. Video URL fields accept a hosted driving video; they do not imply video-upload support.
5. Schema
Input
imagestring
URL of the image used as the first frame (JPEG, PNG, BMP, or WebP). Cannot be combined with `images`, `videos`, or `audios`.
audiosarray
Up to 5 reference audio URLs (WAV or MP3; 1-15s each, 15s total, under 15MB each). Cannot be combined with `image` or `end_image`.
imagesarray
Up to 10 reference image URLs (JPEG, PNG, BMP, or WebP; up to 20MB each, 240-8000px per side). Refer to them in the prompt as Image 1, Image 2, and so on. Cannot be combined with `image` or `end_image`.
promptstringrequired
Text instruction for the generated video.
Default: "A cinematic tracking shot of a sailboat crossing a glowing sea at sunset"
videosarray
Up to 5 reference video URLs (MP4 or MOV; 1-15s each, 15s total, under 100MB each). Cannot be combined with `image` or `end_image`.
durationvalue
Output duration in seconds (2-30), or `auto` to let the model pick a length from the prompt and inputs. `auto` is pre-authorized at 30 seconds and billed on the actual output length.
Default: 5
Possible values: "auto", 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30
end_imagestring
URL of the image used as the last frame (JPEG, PNG, BMP, or WebP). Cannot be combined with `images`, `videos`, or `audios`.
resolutionstring
Output video resolution.
Default: "720p"
Possible values: "480p", "720p", "1080p"
aspect_ratiostring
Output aspect ratio. Omit to let the model choose: 16:9 for text-only requests, otherwise matching the input image.
Possible values: "21:9", "16:9", "4:3", "1:1", "3:4", "9:16"
generate_audioboolean
Generate a synchronized audio track. Pricing is the same either way.
Default: true
{
"model": "alibaba/wan/3.0/prime/video",
"image": "https://static.sandbase.ai/samples/image-to-video/i2v-black-cat-sky-img0.png",
"prompt": "A cinematic tracking shot of a sailboat crossing a glowing sea at sunset",
"duration": 5,
"resolution": "720p",
"generate_audio": true
}Output
A completed run exposes its result in output. The following is the model’s saved example, not a new generation.
No example output has been published yet. Use Playground → Result → JSON after a completed run to inspect your response.