> ## Documentation Index
> Fetch the complete documentation index at: https://docs.jimmyai.cn/llms.txt
> Use this file to discover all available pages before exploring further.
# Fast I2V Create Video
> 创建 Seedance 2.0 视频生成任务。支持满血版（sd2_mx_*，按秒计费，素材须审核）、SP 经济版（seedance2.0-sp / seedance2.0-fast-sp，按分辨率按秒计费）、SP 官方版（seedance2.0-of-sp / seedance2.0-of-fast-sp，按分辨率按秒计费）、Mini 版（seedance2.0-mini，按分辨率按秒计费）、Mini 特价版（seedance2.0-mini-sp，按次计费，仅 480p/720p）、MD（seedance2.0-md / seedance2.0-fast-md，按次计费，720p，不支持音频）、Fast I2V（seedance2.0-fast-i2v，按次计费，720p，仅图片参考最多 9 张，不支持参考视频/音频）、STD 标准版（seedance2.0-std，按次计费，720p，支持音频参考）。创建后使用 GET /api/open-api/v1/videos/{taskId} 查询结果。详见 seedance/md/create、seedance/md/fast-i2v、seedance/md/std、seedance/mini/create 与 seedance/mini-sp/create 文档。
Create a Seedance **Fast I2V** (`seedance2.0-fast-i2v`) video task. Only image references are supported; reference videos and reference audio are not supported. On success you receive a `task_id`. Poll [Query Video Task](/en/api-reference/common/query) until `completed`, then read the video URL from `result.video_url`.
This endpoint is shared with [Create Seedance Full Video](/en/api-reference/seedance/create): `POST /api/open-api/v1/seedance/videos`. Both the request and billing model name is `seedance2.0-fast-i2v`.
## Model
| Route | `model` value | Billing | Duration | Output |
| -------- | ---------------------- | -------- | -------- | ---------- |
| Fast I2V | `seedance2.0-fast-i2v` | Per task | 1–15 s | Fixed 720p |
## Request fields
| Field | Type | Required | Description |
| ------------- | --------- | -------- | ------------------------------------ |
| `model` | string | Yes | Always `seedance2.0-fast-i2v` |
| `prompt` | string | Yes | Text prompt, max **5000** characters |
| `duration` | int | Yes | Length in seconds, 1–15 (default 5) |
| `ratio` | string | No | `16:9` (default), `9:16`, or `1:1` |
| `images` | string\[] | No | Reference images |
| `first_image` | string | No | First frame |
| `last_image` | string | No | Last frame |
`seedance2.0-fast-i2v` does **not** support `reference_videos` or `reference_audios`. The `resolution` field is ignored; output is always 720p.
## Assets
* Pass publicly reachable `https://` image URLs directly.
* SP asset library (`asset://`) and Full-route audit flows do not apply.
## Constraints
* Do not combine `images` (reference mode) with `first_image` / `last_image` (frame mode).
* At most **9** reference images in total (`images` + non-empty `first_image` / `last_image`).
* `ratio` must be one of `16:9`, `9:16`, `1:1`.
* Prompt at most 5000 characters.
## Examples
### Reference images
```bash theme={null}
curl -X POST 'https://www.jimmyai.cn/api/open-api/v1/seedance/videos' \
-H 'Authorization: Bearer sk_xxx' \
-H 'Content-Type: application/json' \
-d '{
"model": "seedance2.0-fast-i2v",
"prompt": "Subject turns slowly, breeze in hair, cinematic lighting",
"duration": 8,
"ratio": "16:9",
"images": [
"https://example.com/ref-1.jpg",
"https://example.com/ref-2.jpg"
]
}'
```
### First and last frame
```bash theme={null}
curl -X POST 'https://www.jimmyai.cn/api/open-api/v1/seedance/videos' \
-H 'Authorization: Bearer sk_xxx' \
-H 'Content-Type: application/json' \
-d '{
"model": "seedance2.0-fast-i2v",
"prompt": "Rainy street at night, girl looks back and smiles, slow dolly in",
"duration": 5,
"ratio": "9:16",
"first_image": "https://example.com/start.png",
"last_image": "https://example.com/end.png"
}'
```
## Response
```json theme={null}
{
"code": 0,
"msg": "ok",
"data": {
"task_id": "vid_abc123",
"status": "queued",
"model": "seedance2.0-fast-i2v",
"created_at": 1710000000
}
}
```
## Poll for result
```bash theme={null}
curl -X GET 'https://www.jimmyai.cn/api/open-api/v1/videos/{task_id}' \
-H 'Authorization: Bearer sk_xxx'
```
Statuses: `queued` → `processing` → `completed` or `failed`. Video URL is in `data.result.video_url` when completed.
Result URLs are retained for about **3 days**. Download promptly.
## OpenAPI
````yaml POST /api/open-api/v1/seedance/videos
openapi: 3.1.0
info:
title: Jimmy AI OpenAPI
description: API for Jimmy AI video generation services
version: 1.0.0
servers:
- url: https://www.jimmyai.cn
description: Production server
security:
- bearerAuth: []
paths:
/api/open-api/v1/seedance/videos:
post:
summary: Create Seedance Video Task
description: >-
Create a Seedance 2.0 video task. Routes: Full/Manxue (sd2_mx_*,
per-second, assets require audit), SP economy (seedance2.0-sp /
seedance2.0-fast-sp, per-second by resolution), SP official
(seedance2.0-of-sp / seedance2.0-of-fast-sp, per-second by resolution),
Mini (seedance2.0-mini, per-second by resolution), Mini Special
(seedance2.0-mini-sp, per-task, 480p/720p only), MD (seedance2.0-md /
seedance2.0-fast-md, per-task, 720p, no audio), Fast I2V
(seedance2.0-fast-i2v, per-task, 720p, image refs only, max 9, no
reference video/audio), STD standard (seedance2.0-std, per-task, 720p,
supports audio references). Poll GET /api/open-api/v1/videos/{taskId}
for results. See seedance/md/create, seedance/md/fast-i2v,
seedance/md/std, seedance/mini/create and seedance/mini-sp/create for
details.
operationId: createSeedanceVideo
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/SeedanceGenerationRequest'
responses:
'200':
description: Task created successfully
content:
application/json:
schema:
$ref: '#/components/schemas/VideoGenerationResponse'
components:
schemas:
SeedanceGenerationRequest:
type: object
required:
- model
- prompt
- duration
properties:
model:
type: string
description: >-
Model name. Manxue: sd2_mx_720p | sd2_mx_1080p | sd2_mx_2k |
sd2_mx_4k; Fast: sd2_mx_fast_720p | sd2_mx_fast_1080p; Video ref:
sd2_mx_video_720p | sd2_mx_video_1080p | sd2_mx_video_2k |
sd2_mx_video_4k; Video ref Fast: sd2_mx_video_fast_720p |
sd2_mx_video_fast_1080p (all per-second). reference_videos only for
video-series models. SP economy: seedance2.0-sp |
seedance2.0-fast-sp (per-second by resolution). SP official:
seedance2.0-of-sp | seedance2.0-of-fast-sp (per-second by
resolution). Mini: seedance2.0-mini (per-second by resolution). Mini
Special: seedance2.0-mini-sp (per-task, 480p/720p only). MD:
seedance2.0-md | seedance2.0-fast-md (per-task, 720p). Fast I2V:
seedance2.0-fast-i2v (per-task, 720p, image refs only). STD
standard: seedance2.0-std (per-task, 720p)
prompt:
type: string
description: 'Text prompt. MD route: max 5000 characters'
duration:
type: integer
description: >-
Duration in seconds. Full 4-12; SP economy 4-15; MD 1-15; STD
standard 4-15. Default 5
resolution:
type: string
enum:
- 480p
- 720p
- 1080p
description: Output resolution for SP economy. Default 720p
orientation:
type: string
enum:
- landscape
- portrait
description: Video orientation (ratio takes precedence if set)
images:
type: array
items:
type: string
description: >-
Reference images. Full route: asset://{assetId}; SP/MD/STD/Fast I2V:
https:// URL. MD max 4; seedance2.0-fast-i2v and STD max 9. Mutually
exclusive with first_image/last_image
ratio:
type: string
description: >-
Aspect ratio. MD: 16:9, 9:16, 1:1 only. STD: 21:9, 16:9, 4:3, 1:1,
3:4, 9:16. Full route also supports 4:3, 3:4, 21:9, Adaptive
first_image:
type: string
description: >-
First frame. Full: asset://{assetId}; Economy/MD: https:// URL. MD
frame mode requires last_image as well; first frame only is not
supported
last_image:
type: string
description: >-
Last frame. Full: asset://{assetId}; Economy/MD: https:// URL. MD
frame mode requires first_image as well
reference_videos:
type: array
items:
type: string
description: >-
Reference videos. Full: asset://{assetId}; Economy/MD: https:// URL.
MD: max 3, combined duration ≤15s, combined size ≤200MB, each
resolution 720px–2160px. Not supported on seedance2.0-fast-i2v
reference_audios:
type: array
items:
type: string
description: >-
Reference audio. Supported on Full, SP economy, and STD. Not
supported on MD or seedance2.0-fast-i2v. Full: asset://{assetId};
SP: https:// URL or asset://. STD max 3 clips; frame mode does not
support audio references
VideoGenerationResponse:
type: object
properties:
code:
type: string
example: '20000'
msg:
type: string
example: ok
data:
type: object
properties:
task_id:
type: string
status:
type: string
model:
type: string
created_at:
type: integer
format: int64
securitySchemes:
bearerAuth:
type: http
scheme: bearer
````