Introduction
Reference-to-video for Veo 3.1 Fast: one to three reference images and a prompt in, an 8 second 720p clip with sound out.
| Field | Value |
|---|---|
| Input | Text |
| Output | Video |
| Publisher | |
| Clip length | 4, 6 or 8 seconds; reference images always give 8 |
| Resolution | 720p: 1280x720 (16:9) or 720x1280 (9:16) |
| Aspect ratios | 16:9, 9:16 |
| Frame rate | 24 fps |
| Sound | Generated with every clip |
| Inputs | Prompt; first frame and optional last frame; 1 to 3 reference images |
Key features & highlights
Put the person from one photo into the place from another
Image 1 was a man in a denim jacket at a gas station, Image 2 an observatory at night. The prompt asked the man from the first image to walk up to the observatory, knock on the door and wave. In 8 seconds he did all three, with the same face, hair and jacket. A woman who was also in Image 2 did not appear.
One continuous push-in
From a single first frame and "Slow push-in", the camera moved in without a cut over 8 seconds, from a wide shot of the dome to a medium close-up. She turned from the telescope, folded her arms and smiled as the stars filled in.
A camera that follows the brief
The same observatory prompt we ran on Veo 3.1 gave a 4-second 16:9 shot down a raised walkway toward the dome. The camera moved forward as asked, and the brass telescope swung into the opening.
Compare tiers before you commit
We sent the same two-sentence paper boat prompt to Veo 3.1 and to Fast. Veo 3.1 filmed from water level with reeds in front; Fast centred the boat on the sun's reflection with ripples spreading round it. Run a few briefs on both before choosing a tier for a project.
Request parameters
| Parameter | Required | Description |
|---|---|---|
prompt | Yes | Required. Name what to take from Image 1, Image 2 and Image 3, then the action and the camera move. at least 1 characters · at most 4000 characters |
reference_image_urls | Yes | Required: 1 to 3 pictures of the people, products or places to keep. In the prompt call them Image 1, Image 2 and Image 3, in the order you upload them. at least 1 items · at most 3 items |
aspect_ratio | No | 16:9 (default) or 9:16. It does not change the price. |
duration_seconds | No | Fixed at 8 seconds in this mode. |
Pricing
Priced as one 8 second clip. How many reference images you send, and the aspect ratio, do not move the price. A job that fails or is declined costs nothing.
| Specification | Price |
|---|---|
| $0.72 / request |
Use cases
Batches of ads and social posts
The price per clip is low enough to run a whole set of 9:16 variants in one sitting: ten 4-second clips cost $3.60, sound included.
The same character or product across several clips
Send 1 to 3 reference images and call them Image 1, Image 2 and Image 3 in the prompt. In our test the face and jacket from Image 1 carried into a new scene.
Storyboards that start and end on set frames
Image to video takes a first frame and an optional last frame, so each shot of a sequence can begin and end where you planned.
Scenes longer than 8 seconds
Continue a clip with video extend: its last frame opens a new 4 to 8 second segment, billed at the image to video price, and the source is not billed.
Prompt guide & pro tips
Refer to references by number
Write "the man from Image 1" and "the observatory from Image 2", in upload order. Say what each one does; the images set the look, the prompt sets the action.
Crop references to what you want
Use one subject per image where you can. Fast left out a second person in our reference image, but other models may bring everything in.
Write the camera and the sound
Name one move, such as "Slow push-in" or "Handheld medium shot", and add a line like "Sound: footsteps on gravel, a metallic knock". Put spoken words in quotes.
Iterate at 4 seconds, finish at 8
At $0.36 a 4-second clip is cheap enough to try several wordings. When one works, render it at 8 seconds, or send the final version to Veo 3.1. Reference clips are 8 seconds from the start, $0.72.
Introduction
Veo 3.1 Fast Reference to video carries a person, a product or a place from up to three images into a new 8 second 720p clip with sound. Google's Veo 3.1 supports reference images itself; on SpicyAPI the mode is offered on Fast and Lite, which makes Fast the Veo tier to reach for when one subject must appear clip after clip. Google screens the prompt and the images first, and a request it turns down fails without a charge.
Other Veo 3.1 Fast modes
- Only have an idea in words? Text to video writes the scene from your prompt alone.
- Starting from a still? Image to video animates a first frame and can finish on a last one.
- Already have footage? Video extend carries a clip on from its last frame.
Prompt tips
- Name the pictures Image 1, Image 2 and Image 3 in the order you upload them.
- Choose clean, well-lit pictures that each show one subject.
- Say what every subject does: the pictures give the look, the prompt gives the action.
API quick start
# Generate the idempotency key once; reuse it when retrying after a lost response.
export SPICY_IDEMPOTENCY_KEY="${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: $SPICY_IDEMPOTENCY_KEY" \
-d '{
"model": "google/veo-3.1-fast/reference-to-video",
"input": {
"aspect_ratio": "16:9",
"duration_seconds": 8,
"prompt": "A slow aerial shot drifting over a misty pine forest at sunrise",
"reference_image_urls": [
"https://example.com/reference.png"
]
}
}'
# Copy data.taskId from the response into TASK_ID, then poll the status.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.spicyapi.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
-H "Authorization: Bearer $SPICY_API_KEY"
See the API reference for authentication, the complete schema, polling, and error handling.