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 Reference to Video
35% OFF
alibaba/wan/3.0/reference-to-videoWan 3.0 by Alibaba generates up to 30-second videos with native audio from reference images, videos, and audio, keeping subjects, motion, and voice consistent with the references.
Try the model
Playground
Input
Text prompt directing how the reference media is used. Reference media can be addressed positionally, e.g. 'the subject in Image 1 walks past Video 1'.
Additional Settings More
Customize your input with more control.
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. Provide at least one of `images`, `videos`, or `audios`.
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 from the prompt and reference inputs. Allowed values: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16.
Output video resolution tier. 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.
ResultExample
What would you like to do next?
Examples
Select an example to load its inputs and result into the playground.
- Open result ↗
Input
{ "aspect_ratio": "9:16", "duration": 15, "images": [ "https://static.sandbase.ai/examples/825e2704-ea2e-446d-bae9-bfaabb6928c9/input/e6ff2dc2-df0d-49cc-870e-198b092f5883.png" ], "prompt": "Using the lip gloss product in Image 1, generate a model swatch video, 15 seconds, vertical screen, highlighting the lip color effect" } - Open result ↗
Input
{ "audios": [ "https://static.sandbase.ai/examples/14ec396a-c777-43c4-acb7-91f7c8cb9d01/input/1f48b8fb-18d8-4431-9355-29adb4604126.mp3" ], "duration": 15, "images": [ "https://static.sandbase.ai/examples/14ec396a-c777-43c4-acb7-91f7c8cb9d01/input/9928ffd8-3ac6-4c82-95eb-f292b818701f.png" ], "prompt": "Seamlessly replace the male character from Image 1 into Video 1's female role. The character sits by the window, holding a phone to the ear. [Voice and dialogue] Extract the voice characteristics from Audio 1, and have the character say the following lines: \"Really? That sounds great! When are you coming back? I miss you so much.\" [Lip-sync and expressions] The generated voice must precisely drive the character's lip movements with natural opening and closing. While speaking, include natural blinking, slight head nodding, and a sense of breathing. Facial micro-expressions must perfectly match the caring and nostalgic emotions in the dialogue. [Lighting and scene] The character's face must naturally receive mixed lighting from warm indoor light and cool neon light from outside the window. Raindrops continuously slide slowly down the window in the background, with slight steam rising from the coffee on the table. The character blends naturally with the scene edges, with no cutout artifacts. [Camera and quality] 15-second long take, camera pushes forward extremely slowly, extremely stable frame, cinematic lighting, 8K resolution, photorealistic quality.", "videos": [ "https://static.sandbase.ai/examples/14ec396a-c777-43c4-acb7-91f7c8cb9d01/input/5b312925-b2ef-40b7-8456-8a70dfe0a374.mp4" ] }
Wan 3.0 Reference to Video
Generate a new shot that reuses what you already have: faces and products from images, motion and framing from clips, voices and music from audio. The prompt decides how the references come together.
Fields
| Field | Type | Default | Notes |
|---|---|---|---|
prompt | string, required | — | Up to 5,000 characters. |
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. |
Send at least one of images, videos, or audios; any combination works.
Billing
You pay per second of output plus, when you send reference videos, their total length (rounded up, counted up to 15s): $0.05 at 480p, $0.10 at 720p, $0.20 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 | $0.25 | $0.50 | $1.00 |
| 10s output | 10s | $0.50 | $1.00 | $2.00 |
| 30s output | 30s | $1.50 | $3.00 | $6.00 |
| 5s output + 8s reference video | 13s | $0.65 | $1.30 | $2.60 |
| 5s output + 15s reference videos | 20s | $1.00 | $2.00 | $4.00 |
With duration: "auto" the run is pre-authorized at 30 seconds, then settled on the length that was actually generated. Failed runs are not charged. Output files are copied to SandBase storage and the returned URL stays valid for 7 days; download anything you need to keep.
Prompting
- Give every reference a job: “Image 1 is the cyclist, Video 1 provides the camera move.”
- Several images of the same subject from different angles hold identity better than one.
- Keep reference audio short and focused on the voice or sound you want carried over.
Image 1 is the cyclist; Video 1 supplies the camera movement. She rides through a rainy city at night in the same red jacket, with quiet traffic in the background.
Limits
- Up to 10 images, 5 videos (15s total), and 5 audio clips (15s total).
- Reference video length plus output length must stay within 30 seconds.
- First/last frames are not accepted here; use image-to-video.
Also on SandBase
Other Wan 3.0 endpoints: alibaba/wan/3.0/video (all modes in one schema), alibaba/wan/3.0/text-to-video, alibaba/wan/3.0/image-to-video, alibaba/wan/3.0/video-edit, alibaba/wan/3.0/video-extend.
Start building
Send your first request
Submit this model through the SandBase Unified Run API.
# The quoted heredoc keeps Unicode and shell metacharacters unchanged.
result=$(curl --fail-with-body --silent \
-X POST "https://api.sandbase.ai/v1/run" \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
-H "Content-Type: application/json" \
--data-binary @- <<'SANDBASE_JSON'
{
"model": "alibaba/wan/3.0/reference-to-video",
"images": [
"https://static.sandbase.ai/examples/bytedance/seedance/2.0/reference-to-video/input_image_safe_ocean_0.jpg"
],
"duration": 5,
"resolution": "720p",
"generate_audio": true
}
SANDBASE_JSON
)
run_id=$(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
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. Provide at least one of `images`, `videos`, or `audios`.
promptstringrequired
Text prompt directing how the reference media is used. Reference media can be addressed positionally, e.g. 'the subject in Image 1 walks past Video 1'.
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
resolutionstring
Output video resolution tier.
Default: "720p"
Possible values: "480p", "720p", "1080p"
aspect_ratiostring
Output aspect ratio. Omit to let the model choose from the prompt and reference inputs.
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/reference-to-video",
"images": [
"https://static.sandbase.ai/examples/bytedance/seedance/2.0/reference-to-video/input_image_safe_ocean_0.jpg"
],
"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.
[
{
"poster": "https://static.sandbase.ai/examples/posters/8520c01a7e49b3bc2c65.jpg",
"source": "Alibaba Cloud",
"source_url": "https://www.alibabacloud.com/help/en/model-studio/wan3-video-generation-guide",
"url": "https://static.sandbase.ai/examples/825e2704-ea2e-446d-bae9-bfaabb6928c9/output/157c65c0-77e9-4ceb-a031-3bed2d561d11.mp4"
}
]