Seedance 2.5 Economy
Seedance 2.5 Economy is engineered as a high-value supplementary pipeline for large-scale cost reduction, it maintains exceptional motion stability and native audio synchronization whenever connected. Due to tight compute availability dynamically fluctuating around 60%–70% (with optimal daily throughput windows spanning roughly 6–7 hours), programmatical API integration with automated retry and polling logic is strongly recommended. Seedance 2.5 Economy efficiently powers high-volume e-commerce product videos, automated social media content matrices, and low-cost ad creative pipelines.
Seedance 2.5 Economy
Seedance 2.5 Economy is a video generation model focused on generation efficiency and cost control. It can generate videos from 4 to 30 seconds based on text prompts and supports multiple aspect ratios, as well as image, video, and audio references.
When reference videos are used, their duration is included in the billable duration. The total cost is calculated as “unit price × (output video duration + reference video duration).”
Base URL
https://api.icreat.aiAuthentication
All API requests require authentication using an API key. You can obtain an API key from the console.
export ICREAT_API_KEY="your-api-key-here"HTTP Headers
import os
API_KEY = os.environ.get("ICREAT_API_KEY")
headers = {
"Content-Type": "application/json",
"Authorization": "Bearer " + API_KEY,
}Keep Your API Key Secure
Never expose your API key in client-side code or public repositories. Use environment variables or a backend proxy.
Code Examples
Video generation uses a two-step asynchronous workflow: first submit a task and obtain a task_id, then poll the task result endpoint for progress.
The response contains status and result. While the task is processing, result is []. When status is SUCCEEDED, result contains the generated resource list and the response includes costUSD. The same task_id must be used for both steps.
1. Submit a Task
Send a video generation request to the task submission endpoint. Use @Image1, @Video1, and @Audio1 in the prompt to reference the first resource in the corresponding array.
2. Query Task Result (Polling)
Use the task_id returned when submitting the task to query its progress. You can call this endpoint repeatedly until the task reaches a terminal state.
The response contains status and result. While the task is processing, result is []. When status is SUCCEEDED, result contains the generated resource array and the response includes costUSD. When status is FAILED, the task has failed.
Input Parameters
Submit Task - Input Parameters
The task submission request body supports the following parameters.
Total: 8; required: 4; optional: 4
The text prompt describing the video to be generated. Supports multiline text.
When using reference resources, reference the corresponding array items with @ImageN, @VideoN, or @AudioN, where N starts at 1. For example, @Image1 refers to the first image URL in the images array.
A list of publicly accessible image URLs used as image references. Reference them in the prompt using markers such as @Image1 and @Image2.
A list of publicly accessible video URLs used as video references. Reference them in the prompt using markers such as @Video1 and @Video2.
The duration of reference videos is included in the billable duration. The total cost is calculated as “unit price × (output video duration + reference video duration).”
A list of publicly accessible audio URLs used as audio references. Reference them in the prompt using markers such as @Audio1 and @Audio2.
The output video resolution.
The aspect ratio of the output video. When set to adaptive, the model automatically selects a suitable aspect ratio based on the input content.
The output video duration in seconds. Accepts integers from 4 to 30, or -1 to let the model select the duration automatically.
Min: 4 Max: 30
Whether to add an “AI Generated” watermark to the output video.
Query Task Result - Input Parameters
Total: 1; required: 1; optional: 0
The task ID returned by the task submission endpoint.
Output Parameters
Submit Task - Output Parameters
Total: 1
The task ID returned by the task submission endpoint.
Query Task Result - Output Parameters
Total: Variable
The current task status. Before the task succeeds, result is usually []. When the status is SUCCEEDED, result contains the generated resources and the response includes costUSD.
The list of generated resources. This field is [] while the task is processing or has not succeeded. After the task succeeds, it contains an array of resource objects.
The task cost in US dollars. This field is returned only when status is SUCCEEDED.
LLM-Friendly Prompt
Billing
- The total cost is calculated as “unit price × (output video duration + reference video duration).”
- When no reference video is provided, only the output video duration is billed.
- When reference videos are provided, the duration of all billable reference videos is included in the total billable duration.
- Image and audio references are not included in the video duration.
The following is an LLM-friendly Markdown prompt that can be copied into Cursor, ChatGPT, or other AI assistants to help them understand this model’s API integration, workflow, and key parameters. Click the “Copy LLM Prompt” button or copy the complete content from the code block below.
# bytedance/seedance-2-5-economy
> Seedance 2.5 Economy is a video generation model focused on generation efficiency and cost control. It supports text prompts as well as image, video, and audio references.
## Overview
Use iCreat’s two-step asynchronous task API to submit a video generation request, then poll the task result endpoint. After the task succeeds, retrieve the video resource URL from `result`. The successful response also includes `costUSD`.
When reference videos are used, their duration is included in the billable duration. The total cost is calculated as “unit price × (output video duration + reference video duration).”
## API Information
- **Base URL**: `https://api.icreat.ai`
- **Task submission endpoint (POST)**: `/v1/task/submit/bytedance/seedance-2-5-economy`
- **Task result endpoint (POST)**: `/v1/task/result`
- **Model ID**: `bytedance/seedance-2-5-economy`
- **Authentication**: `Authorization: Bearer ${ICREAT_API_KEY}`
## Workflow
1. **Submit a task**: Send a POST request to the task submission endpoint with the generation parameters in the request body. The endpoint returns `{ "task_id": "..." }`.
2. **Query the result**: Send a POST request to `/v1/task/result` with `{ "task_id": "..." }` in the request body.
3. While the task is processing, `result` is `[]`. Continue polling.
4. When the status is `SUCCEEDED`, read `url` or `download_url` from the resource whose `type` is `Video`.
5. When the status is `FAILED`, stop polling and check the request parameters and reference resources.
## Task Submission Example
```json
{
"prompt": "Use the subject from @Image1, the camera movement from @Video1, and the rhythm from @Audio1 to generate a video",
"images": [
"https://example.com/reference.png"
],
"videos": [
"https://example.com/reference.mp4"
],
"audios": [
"https://example.com/reference.mp3"
],
"resolution": "480p",
"ratio": "9:16",
"duration": 4,
"watermark": true
}
```
## Input Parameters
- `prompt` (required): The video generation prompt. Supports multiline text. Use `@ImageN`, `@VideoN`, and `@AudioN` to reference resources in the corresponding arrays. Indexing starts at `1`.
- `images` (optional): A list of publicly accessible image URLs. Reference them using markers such as `@Image1` and `@Image2`.
- `videos` (optional): A list of publicly accessible video URLs. Reference them using markers such as `@Video1` and `@Video2`. Reference video duration is included in the billable duration.
- `audios` (optional): A list of publicly accessible audio URLs. Reference them using markers such as `@Audio1` and `@Audio2`.
- `resolution` (required): The output video resolution. Supported values are `480p`, `720p`, and `1080p`.
- `ratio` (required): The output video aspect ratio. Supported values are `16:9`, `4:3`, `1:1`, `3:4`, `9:16`, `21:9`, and `adaptive`.
- `duration` (required): The output video duration. Accepts integers from `4` to `30`, or `-1` to let the model select the duration automatically. The default value is `5`.
- `watermark` (optional): Whether to add an “AI Generated” watermark to the output video.
## Billing
- Total cost = unit price × (output video duration + reference video duration).
- When no reference video is provided, only the output video duration is billed.
- Image and audio references are not included in the video duration.
## Output
- Check `status` when polling the task. While the task is processing, `result` is `[]`.
- After the task succeeds, the response has the format `result: [{ "type": "Video", "url": "...", "download_url": "..." }]` and includes `costUSD`.
## Notes
- The same `task_id` must be used when submitting the task and querying its result.
- `FAILED` is a terminal state. When a task fails, check the request parameters and reference resources.
- The order of the reference resource arrays must match the reference indexes used in the prompt.