Seedance 2.0 Fast Image-to-Video API

Overview

Seedance 2.0 Fast Image-to-Video is built for the moment after a visual has already been approved but the motion still needs revision. Keep the same subject, composition, product design, or key art, then test a different camera path, gesture, pace, environmental movement, sound cue, or ending.

The endpoint uses one required First Frame and one required Prompt. Through the API, an optional Last Frame can define where the shot should finish. The model supports 480P and 720P output, durations from 4 to 15 seconds or automatic duration, fixed or adaptive aspect ratios, and optional generated audio.

Fast is not the lowest-cost Seedance tier and it is not the highest-resolution tier. Choose it when shortening the motion-feedback loop matters more than minimizing batch cost or producing a 1080P or 4K final asset.

Use the model through:

bytedance/seedance-2-0-fast/image-to-video

Key Features

The approved image remains the fixed creative starting point. The First Frame preserves the intended opening subject, composition, materials, color treatment, and lighting while the motion brief changes.

One revision can isolate one motion variable. Keep the same image and parameters while testing a new camera move, gesture, pacing decision, environmental effect, or audio cue.

An optional Last Frame can test the ending. Through the API, a Last Frame gives the shot a destination without requiring the opening visual to be redesigned.

480P and 720P support different review stages. Use 480P to judge motion structure quickly and 720P when facial detail, product shape, timing, or transition quality needs closer review.

Audio can be revised with the movement. Set generate_audio to true to compare dialogue, ambience, sound effects, or music while keeping the visual source unchanged.

Fixed settings support fair A/B tests. Use the same ratio, resolution, and duration when comparing two motion directions. Use adaptive or duration: -1 only when exploration matters more than direct comparison.

Protected images require the correct review setting. Set need_review: true when an image contains faces or copyrighted IP. Incorrect settings can cause failure or increase processing time.

Model Comparison

Seedance 2.0 Mini vs Fast vs Mainline Image-to-Video

Seedance 2.0 Fast I2V vs Wan 2.7 I2V SP vs HappyHorse 1.1 I2V Spicy

Inputs

The Playground accepts one starting image. The API also supports an optional Last Frame for a controlled destination.

Each image must have an aspect ratio from 0.4 to 2.5, dimensions from 300 to 6,000 pixels, and a file size under 30 MB. The total request body must remain under 64 MB.

Chinese Prompts can contain up to 2,000 characters, and English Prompts can contain up to 2,000 words. For revision work, describe the requested change and the details that must stay fixed instead of rewriting the entire source image.

Set need_review: true when a submitted image contains faces or copyrighted IP.

Parameters

Pricing

Total Cost = Unit Price × Output Video Duration

Quick Start

This request keeps the approved product image but changes the camera movement and ending for a new review version.

curl --fail-with-body --connect-timeout 10 --max-time 60 \
  -X POST https://api.icreat.ai/v1/task/submit/bytedance/seedance-2-0-fast/image-to-video \
  -H "Authorization: Bearer ${ICREAT_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "content": [
      {
        "type": "text",
        "text": "Keep the approved product shape, materials, colors, logo placement, lighting style, and opening composition. Replace the previous static hold with one smooth clockwise camera orbit while the product remains centered. Add a short light sweep across the front surface at the midpoint, then slow the camera before reaching the supplied last frame. Keep the movement controlled and avoid changing the background. Add a soft mechanical rise and one restrained impact sound at the ending."
      },
      {
        "type": "image_url",
        "image_url": {
          "url": "https://example.com/approved-first-frame.jpg"
        },
        "role": "first_frame",
        "need_review": false
      },
      {
        "type": "image_url",
        "image_url": {
          "url": "https://example.com/revised-last-frame.jpg"
        },
        "role": "last_frame",
        "need_review": false
      }
    ],
    "generate_audio": true,
    "ratio": "16:9",
    "resolution": "720p",
    "duration": 7,
    "watermark": false
  }'

The request returns a task_id.

Submit request → Receive task_id → Poll status → Retrieve result

Query the task status until it reaches SUCCEEDED, then use the same task_id to retrieve the result.

Output

Seedance 2.0 Fast Image-to-Video uses an asynchronous task flow.

The submission request returns a task_id. Task status is checked through the query-status endpoint, and the completed result is retrieved through the get-result endpoint after the task succeeds.

Detailed result fields are not listed until they are confirmed from an actual iCreat response.

Use Cases

Stakeholder motion feedback on approved key art. Change the gesture, camera energy, environmental movement, or ending after the visual design has already been accepted.

Product-camera A/B tests. Compare a push-in, orbit, tilt, or locked-camera treatment around the same hero image without changing the product composition.

Character-performance revisions. Test a calmer expression, faster reaction, different hair movement, or revised clothing motion while the character design remains fixed.

First-to-last transition rehearsal. Check whether two approved frames connect naturally before moving the selected version to a higher-resolution endpoint.

Short-form pacing revisions. Compare a faster reveal, longer hold, earlier sound cue, or different final beat for one social asset.

Limitations

Fast is limited to 480P and 720P. It is designed for motion review and revision, not as a replacement for the 1080P or 4K output tiers of the mainline Seedance 2.0 Image-to-Video endpoint.

The source image is a creative anchor, not a pixel-locked frame. Faces, hands, product geometry, labels, textures, lighting, and small background details may change when the model creates unseen movement.

Large rotations, rapid camera travel, major pose changes, and strong perspective shifts require the model to invent more information beyond the source image. These changes increase the risk of drift, flicker, or altered proportions.

A Last Frame can make the ending more controllable, but incompatible First and Last Frames can destabilize the transition. The subject, viewpoint, scale, lighting, and environment should form a believable path.

Changing the output ratio may crop, extend, or reposition the source composition. For fair A/B tests, keep ratio, resolution, duration, and generated-audio settings unchanged.

Readable text and logos remain unreliable during motion. Generated audio may contain unclear dialogue, distortion, or timing differences. The endpoint does not expose a seed, so exact repeatability should not be expected.

FAQ

When should Fast I2V be chosen instead of Mini or Mainline I2V?

Choose Fast when the approved image needs several motion revisions and turnaround matters more than the lowest batch cost. Choose Mini for the lowest-cost batch variants and Mainline for selected 1080P, 4K, or more demanding final output.

Can Fast Image-to-Video use a Last Frame?

Yes. The API supports an optional last_frame that defines the intended ending. The current Playground shows one image upload, so Last Frame control is an API capability.

How can two motion revisions be compared fairly?

Use the same First Frame, ratio, resolution, duration, and audio setting. Change only one motion variable, such as the camera path, gesture, pace, environmental effect, or ending.

Can several reference images or videos be used on this endpoint?

Not in this Image-to-Video workflow. Use the Fast Reference-to-Video endpoint when several images, videos, or audio files must guide the result.

Should review runs use 480P or 720P?

Use 480P for early motion structure and broad iteration. Use 720P when facial detail, product shape, transition quality, or audio timing needs closer review.