> ## Documentation Index
> Fetch the complete documentation index at: https://docs.apiyi.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Wan2.7 Reference-to-Video API 참조
> Wan2.7-r2v reference-to-video API 참조 및 라이브 플레이그라운드: 참조 이미지/동영상의 주제 특징을 유지하며, 다중 주제 상호작용, 음성 참조, 분할 화면 스토리보드를 지원합니다.
오른쪽의 Playground에서는 직접 디버깅할 수 있습니다: `Bearer sk-your-api-key`를 **Authorization**에 넣고, `model` / `input.media` / `parameters`를 채운 다음 요청을 보내십시오. 성공적으로 제출하면 `task_id`가 반환됩니다. 폴링 및 다운로드는 아래를 참조하십시오.
이 페이지는 `wan2.7-r2v`(reference-to-video)의 생성 엔드포인트입니다. 참조 이미지/비디오를 제공하면 모델이 대상(사람/동물/사물)과 장면 특징을 유지하면서, 단일 캐릭터 연기 또는 다중 캐릭터 상호작용을 생성합니다. 더 많은 참조 이미지가 필요하면(최대 9개), [HappyHorse r2v](/ko/api-capabilities/happyhorse/reference-to-video)를 고려해 보십시오. 전체 비동기 흐름은 [Wan 개요](/ko/api-capabilities/wan/overview)를 참조하십시오.
* **참조 자산 인용 규칙**: prompt에서 "image 1 / image 2"는 `reference_image`를, "video 1 / video 2"는 `reference_video`를 가리키는 데 사용하며, `media` 배열과 같은 순서로 작성하십시오(이미지와 비디오는 각각 별도로 계산합니다). 이미지/비디오가 하나뿐이면 "the reference image" / "the reference video"라고 간단히 쓰면 됩니다.
* **개수 제한**: `reference_image` + `reference_video` 총합 ≤5; 최대 `first_frame` 1개입니다.
* 생성 요청은 `/wan/api/v1/...`에 `X-DashScope-Async: enable`를 사용해 보내십시오; `/v1/videos`는 사용하지 마십시오.
## 코드 예시
```bash cURL theme={null}
curl -X POST "https://api.apiyi.com/wan/api/v1/services/aigc/video-generation/video-synthesis" \
-H "X-DashScope-Async: enable" \
-H "Authorization: Bearer sk-your-api-key" \
-H "Content-Type: application/json" \
-d '{
"model": "wan2.7-r2v",
"input": {
"prompt": "A girl wearing this gown walks slowly through a garden bathed in sunset, the breeze gently lifting her skirt, cinematic lighting",
"media": [
{"type": "reference_image", "url": "https://your-cdn.com/dress.png"}
]
},
"parameters": {"resolution": "720P", "duration": 5, "prompt_extend": true, "watermark": true}
}'
```
```python Python (단일 이미지 참조) theme={null}
import requests
url = "https://api.apiyi.com/wan/api/v1/services/aigc/video-generation/video-synthesis"
headers = {
"Authorization": "Bearer sk-your-api-key",
"Content-Type": "application/json",
"X-DashScope-Async": "enable",
}
body = {
"model": "wan2.7-r2v",
"input": {
"prompt": "The reference image: a girl walking slowly through a garden bathed in sunset, cinematic lighting",
"media": [{"type": "reference_image", "url": "https://your-cdn.com/girl.png"}],
},
"parameters": {"resolution": "720P", "duration": 5, "prompt_extend": True},
}
resp = requests.post(url, json=body, headers=headers, timeout=30)
print("task_id:", resp.json()["output"]["task_id"])
```
```python Python (다중 피사체 + 음성) theme={null}
import requests
# Multi-subject: image 1 = girl (with voice), video 1 = boy (with voice), image 2/3 = props/background
body = {
"model": "wan2.7-r2v",
"input": {
"prompt": "Video 1 holds image 2, walks past image 1, and says: the sunshine is lovely today.",
"media": [
{"type": "reference_image", "url": "https://your-cdn.com/girl.jpg",
"reference_voice": "https://your-cdn.com/girl-voice.mp3"},
{"type": "reference_video", "url": "https://your-cdn.com/boy.mp4",
"reference_voice": "https://your-cdn.com/boy-voice.mp3"},
{"type": "reference_image", "url": "https://your-cdn.com/object.png"},
],
},
"parameters": {"resolution": "720P", "ratio": "16:9", "duration": 10, "prompt_extend": False, "watermark": True},
}
resp = requests.post(
"https://api.apiyi.com/wan/api/v1/services/aigc/video-generation/video-synthesis",
json=body,
headers={"Authorization": "Bearer sk-your-api-key", "Content-Type": "application/json",
"X-DashScope-Async": "enable"},
timeout=30,
)
print(resp.json()["output"]["task_id"])
```
## 매개변수 및 미디어 빠른 참조
| 매개변수 | 유형 | 필수 | 기본값 | 비고 |
| -------------------------- | --- | -- | ------- | ----------------------------------------------------------- |
| `model` | 문자열 | ✓ | — | `wan2.7-r2v`으로 고정 |
| `input.prompt` | 문자열 | ✓ | — | ≤5000자; 자산을 참조하려면 "image 1/video 1"을 사용합니다 |
| `input.media` | 배열 | ✓ | — | 아래 미디어 표를 참조하십시오 |
| `parameters.resolution` | 문자열 | | `1080P` | `720P` / `1080P` |
| `parameters.ratio` | 문자열 | | `16:9` | `16:9` / `9:16` / `1:1` / `4:3` / `3:4` (첫 프레임이 전달되면 무시됩니다) |
| `parameters.duration` | 정수 | | `5` | 참조 비디오가 있으면 2-10, 없으면 2-15 |
| `parameters.prompt_extend` | 불리언 | | `true` | 스마트 재작성 |
| `parameters.watermark` | 불리언 | | `false` | “AI 생성” 워터마크 |
### `media[]` 값
| `type` | 개수 / 제한 | 비고 |
| ----------------- | ------------- | ------------------------------------------------------------------------------ |
| `reference_image` | 비디오와 함께 최대 5개 | 참조 이미지입니다. 피사체(사람/동물/물체) 또는 장면을 제공합니다. 음성을 설정하려면 `reference_voice`를 첨부할 수 있습니다 |
| `reference_video` | 이미지와 함께 최대 5개 | 참조 비디오입니다. 피사체와 음성 참조를 제공합니다. 빈 장면 영상은 전달하지 마십시오 |
| `first_frame` | 최대 1개 | 선택 사항인 첫 프레임입니다. 시작 프레임을 함께 제어합니다 |
| `reference_voice` | 첨부된 필드 | `reference_image`/`reference_video`에 첨부하여 해당 피사체의 음성을 설정합니다(wav/mp3, 1-10초) |
## 응답 형식
```json theme={null}
{
"output": { "task_id": "acda59b4-3b10-4789-a5e5-edadae48adcb", "task_status": "PENDING" },
"request_id": "..."
}
```
## 상태 확인 및 다운로드
`task_id`를 받으면, mp4의 상태를 조회하고 다운로드하기 위해 다음 세 단계를 따르십시오.
* **조회** `GET /v1/tasks/{task_id}` (`Authorization` 포함; 쿼리는 `X-DashScope-Async` 헤더가 **필요하지 않습니다**), 10초마다(3초 미만은 안 됨) `status`가 `completed`가 될 때까지 반복하십시오. Reference-to-video는 보통 1\~5분이 걸리므로, 대체 수단으로 20분의 클라이언트 타임아웃을 설정하십시오.
* **상태 값**: `submitted`(대기 중) / `in_progress`(생성 중; `progress`가 30%에서 멈춰 있는 것은 정상입니다) / `completed`(성공) / `failed`(`error` 확인).
* **다운로드**: 응답에서 `result_url`를 직접 GET 하며, **`Authorization` 헤더 없이** 수행하십시오(이는 OSS 서명된 직접 링크입니다. Auth를 보내면 403이 반환됩니다); `result_url`는 기본적으로 **24시간** 후 만료되므로, 즉시 저장하십시오.
```bash Check status theme={null}
curl "https://api.apiyi.com/v1/tasks/acda59b4-3b10-4789-a5e5-edadae48adcb" \
-H "Authorization: Bearer sk-your-api-key"
```
```json Completed response theme={null}
{
"status": "completed",
"progress": 100,
"result_url": "https://dashscope-result-xxx.oss-cn-beijing.aliyuncs.com/xxx.mp4?Expires=...&Signature=...",
"task_id": "acda59b4-3b10-4789-a5e5-edadae48adcb"
}
```
```bash Download video theme={null}
# result_url is an OSS signed direct link — do NOT send the Authorization header (it returns 403)
curl -L -o out.mp4 "https://dashscope-result-xxx.oss-cn-beijing.aliyuncs.com/xxx.mp4?Expires=...&Signature=..."
```
```bash One command: check + download theme={null}
TASK_ID="acda59b4-3b10-4789-a5e5-edadae48adcb"
URL=$(curl -s "https://api.apiyi.com/v1/tasks/$TASK_ID" \
-H "Authorization: Bearer sk-your-api-key" | jq -r '.result_url')
curl -L -o out.mp4 "$URL" # no Authorization
```
위 내용은 “이미 task\_id가 있는 경우”를 위한 빠른 조회/다운로드 경로입니다. 전체 폴링 루프(타임아웃 대체 수단 포함)와 Python 클라이언트는 [Wan 개요 · 비동기 호출 흐름](/ko/api-capabilities/wan/overview#async-call-flow)을 참조하십시오.
## OpenAPI
````yaml api-reference/wan-reference-to-video-openapi-en.yaml POST /wan/api/v1/services/aigc/video-generation/video-synthesis
openapi: 3.1.0
info:
title: Wan2.7 Reference-to-Video API
description: >
Alibaba Cloud Tongyi Wanxiang Wan2.7 reference-to-video (`wan2.7-r2v`) —
preserve subject features from reference images/videos, with multi-subject
interaction, voice reference, and split-screen storyboards. DashScope async
passthrough endpoint.
- Creation requests must include the header `X-DashScope-Async: enable`
- In the prompt, use "image 1 / video 1" to refer to reference assets, in
the same order as the media array
- `reference_image` + `reference_video` total <=5; `first_frame` <=1
- Async task endpoint: returns a `task_id`; poll `GET /v1/tasks/{task_id}`,
then download from `result_url` (no Authorization header, expires in 24
hours)
- **Do not use `/v1/videos`**
**Authentication**: add `Authorization: Bearer YOUR_API_KEY` to the request
header
**Get an API Key**: visit the APIYI console `api.apiyi.com/token` to create
a Token
version: 1.0.0
servers:
- url: https://api.apiyi.com
description: Primary endpoint
- url: https://vip.apiyi.com
description: Backup endpoint
security:
- bearerAuth: []
paths:
/wan/api/v1/services/aigc/video-generation/video-synthesis:
post:
tags:
- Video Generation
summary: >-
Reference-to-video: create a video task that preserves subject features
from reference images/videos
description: >
Submit a `wan2.7-r2v` reference-to-video task (async); returns a
`task_id`.
- Required: `model`, `input.prompt`, `input.media` (at least 1
reference_image / reference_video), header `X-DashScope-Async: enable`
- Optional: `reference_voice` (voice reference), `parameters` (including
ratio)
- **The response does not contain the video file**; poll `GET
/v1/tasks/{task_id}` until `completed`, then download from `result_url`
- 720P / 5 seconds typically takes 120-140 seconds; multi-subject tasks
may take 1-5 minutes
operationId: createWan27ReferenceToVideo
parameters:
- name: X-DashScope-Async
in: header
required: true
description: Async processing switch; must be set to enable
schema:
type: string
enum:
- enable
default: enable
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/WanReferenceToVideoRequest'
example:
model: wan2.7-r2v
input:
prompt: >-
A girl wearing this gown walks slowly through a garden bathed
in sunset, the breeze gently lifting her skirt, cinematic
lighting
media:
- type: reference_image
url: https://your-cdn.com/dress.png
parameters:
resolution: 720P
ratio: '16:9'
duration: 5
prompt_extend: true
watermark: true
responses:
'200':
description: Task submitted; returns task_id and PENDING status
content:
application/json:
schema:
$ref: '#/components/schemas/WanVideoTask'
'400':
description: Invalid parameter or more than 5 reference assets
'401':
description: Unauthorized - invalid API Key
'403':
description: Blocked by content moderation, or no permission for the group
'429':
description: Rate limit exceeded or insufficient balance
'500':
description: Missing media or upstream error; check the body and retry
security:
- bearerAuth: []
components:
schemas:
WanReferenceToVideoRequest:
type: object
required:
- model
- input
properties:
model:
type: string
description: >-
Model ID; for reference-to-video use wan2.7-r2v (legacy versions
wan2.6-r2v / wan2.6-r2v-flash)
enum:
- wan2.7-r2v
- wan2.6-r2v
- wan2.6-r2v-flash
default: wan2.7-r2v
input:
type: object
required:
- prompt
- media
properties:
prompt:
type: string
description: >-
Text prompt, <=5000 characters; use 'image 1/video 1' to refer
to reference assets
example: >-
The reference image: a girl walking slowly through a garden
bathed in sunset, cinematic lighting
negative_prompt:
type: string
description: Negative prompt, <=500 characters
media:
type: array
description: >-
Media asset array; reference_image + reference_video total <=5,
first_frame <=1
items:
type: object
required:
- type
- url
properties:
type:
type: string
description: >-
Media type: reference_image (reference image) /
reference_video (reference video) / first_frame (optional
first frame)
enum:
- reference_image
- reference_video
- first_frame
url:
type: string
description: Public https link fetchable directly with GET
example: https://your-cdn.com/dress.png
reference_voice:
type: string
description: >-
Optional: voice reference audio URL (wav/mp3, 1-10s) to
set this subject's voice
parameters:
$ref: '#/components/schemas/WanParameters'
WanVideoTask:
type: object
properties:
output:
type: object
properties:
task_id:
type: string
description: >-
Task ID, used to poll GET /v1/tasks/{task_id}, valid for 24
hours
example: acda59b4-3b10-4789-a5e5-edadae48adcb
task_status:
type: string
enum:
- PENDING
- RUNNING
- SUCCEEDED
- FAILED
example: PENDING
request_id:
type: string
description: Unique request identifier
example: ...
WanParameters:
type: object
properties:
resolution:
type: string
description: >-
Resolution tier (uppercase); wan2.7-r2v supports 720P / 1080P,
default 1080P
enum:
- 720P
- 1080P
default: 1080P
ratio:
type: string
description: Aspect ratio; ignored automatically when a first_frame is supplied
enum:
- '16:9'
- '9:16'
- '1:1'
- '4:3'
- '3:4'
default: '16:9'
duration:
type: integer
description: >-
Video duration (seconds), integer; range 2-10 with a reference
video, 2-15 without
minimum: 2
maximum: 15
default: 5
prompt_extend:
type: boolean
description: Whether to enable smart prompt rewriting
default: true
watermark:
type: boolean
description: Whether to add an AI generated watermark in the bottom-right corner
default: false
seed:
type: integer
description: Random seed, 0-2147483647
minimum: 0
maximum: 2147483647
securitySchemes:
bearerAuth:
type: http
scheme: bearer
description: The API Key obtained from the APIYI console
````