Vidu（Reference-based video generation）

POST

/vidu/ent/v1/tasks

A rising video generation company from China, whose latest 1.5 model features an exclusive subject generation capability.
Relevant link：https://www.vidu.studio/

Model Version	Function	Duration (seconds)	Resolution	Style	Generation Time (seconds)	Price (ptc)
Vidu-1.5	Reference-based video generation (character2video)	4	360p	General	30	0.5
		4	720p	General	155	1.25
		4	1080p	General	620	2.5
		8	720p	General	360	2.5
Vidu-1.0	Reference-based video generation (character2video)	4	360p	General	30	0.5
		8	360p	General	60	1.0

Price: Please calculate based on credits, 1 credit = 0.05 PTC
Official website pricing link: https://platform.vidu.cn/pricing

Request

Header Params

Authorization

string

optional

Example:

Bearer {{YOUR_API_KEY}}

Content-Type

string

optional

Example:

application/json

Body Params application/json

type

enum<string>

required

Allowed values:

text2videoimg2videocharacter2videoheadtailimgupscale

model

string

optional

Model types: text2video, img2video, and
character2video

Available values for tasks:

vidu-high-performance (for the above tasks)

stable (for the upscale task).

When the model version is set to 1.0, both vidu-high-performance and stable must be provided. When the version is set to 1.5, these parameters are optional. These two parameters do not affect the generation quality and only need to be transmitted as required.

model_version

enum<string>

required

Allowed values:

1.01.5

style

enum<string>

optional

Style: Required only for the text2video task.
Options: general,anime.

Allowed values:

generalanime

moderation

boolean

optional

Audit toggle: Specifies whether to enable content review.
Default value: true

input

object

required

The task input includes the following fields

seed

integer

optional

Random seed, if not provided or if 0 is passed, a random number will be used instead

multi_image_boost

boolean

optional

Enable multi-object reference

enhance

boolean

optional

Whether to use the prompt enhancement feature, default value is true

creation_id

string

optional

Generation ID, required only for the upscale task

prompts

array [object {2}]

optional

output_params

object

optional

sample_count

integer

required

Number of videos to generate, currently only supports 1

duration

integer

required

Video duration, supports 4 or 8

resolution

enum<string>

required

Resolution, only supported by model 1.5. Option: 512. This means the base resolution is 512x512

Allowed values:

5127201080

movement_amplitude

enum<string>

optional

Motion range, only supported by model 1.5. Options: auto, small, medium, large. If auto is selected, the model will automatically choose the most suitable motion range based on the input

Allowed values:

autosmallmediumlarge

aspect_ratio

string

required

Video aspect ratio, only supported by model 1.5. Options: default, 16:9, 9:16, 1:1

Example

{
    "type": "character2video",
    "model_version": "1.5",
    "input": {
        "enhance": true,
        "multi_image_boost": true,
        "prompts": [
            {
                "type": "text",
                "content": "女孩的手心漂浮着一只半透明金鱼，镜头逐渐拉近，镜头顺时针旋转"
            },
            {
                "type": "image",
                "content": "https://dev-ss-images.s3.cn-northwest1.amazonaws.com.cn/tmp/20241126-175112.png"
            },
            {
                "type": "image",
                "content": "https://dev-ss-images.s3.cn-northwest1.amazonaws.com.cn/tmp/20241126-175116.png"
            },
            {
                "type": "image",
                "content": "https://dev-ss-images.s3.cn-northwest1.amazonaws.com.cn/tmp/20241126-175120.png"
            }
        ]
    },
    "output_params": {
        "sample_count": 1,
        "resolution": "720p",
        "duration": 4,
        "aspect_ratio": "1:1",
        "movement_amplitude": "auto"
    },
    "moderation": false
}

Request samples

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

curl --location --request POST 'https://api.302.ai/vidu/ent/v1/tasks' \
--header 'Authorization: Bearer sk-jls4AaVBGoe1GwZD64qZA1qyKTN1MPHa4NmvH1cT68z7K1Zz' \
--header 'Content-Type: application/json' \
--data-raw '{
    "type": "character2video",
    "model_version": "1.5",
    "input": {
        "enhance": true,
        "multi_image_boost": true,
        "prompts": [
            {
                "type": "text",
                "content": "女孩的手心漂浮着一只半透明金鱼，镜头逐渐拉近，镜头顺时针旋转"
            },
            {
                "type": "image",
                "content": "https://dev-ss-images.s3.cn-northwest1.amazonaws.com.cn/tmp/20241126-175112.png"
            },
            {
                "type": "image",
                "content": "https://dev-ss-images.s3.cn-northwest1.amazonaws.com.cn/tmp/20241126-175116.png"
            },
            {
                "type": "image",
                "content": "https://dev-ss-images.s3.cn-northwest1.amazonaws.com.cn/tmp/20241126-175120.png"
            }
        ]
    },
    "output_params": {
        "sample_count": 1,
        "resolution": "720p",
        "duration": 4,
        "aspect_ratio": "1:1",
        "movement_amplitude": "auto"
    },
    "moderation": false
}'

Responses

🟢200Success

application/json

Body

object {0}

Example

{}

Modified at 2025-05-13 03:18:05

Vidu（Generate video from the first and last frames）

Vidu（Generate scene video）