Sign in

v1

latestOpenAPI 3.0.02026-07-248574216.0 KB
Text2Video & Image2Video

Text2Video

The Text2Video service helps generate a video based on the text introduced as input by the user. Considering the time and cost to generate videos, only one video will be generated per request (unlike the image generation services). Note, that considering the big number of different models that are supported with this service, some limitations may apply. For example, some models (e.g. Grok Imagine Video / OVI) don't support the option to disable sound. With these models the result video will always be generated regardless to the provided input params. Similarly, note that different models do support different resolutions and output video quality parameters. In this case as well, the output video's resolutions and quality are not guaranteed to be exactly the same as provided input params. The service will provide what's "closer" to the original request. To make sure there are minimal deviations, we recommend doing test requests and checking the output video as well as doing research on original model's official documentation.

post/text2video

Request body

promptstring required

The prompt.

widthnumber

Desired width used to determine the aspect ratio and resolution tier. The actual output width is determined by the AI provider based on the closest matching supported aspect ratio. Supported ratios vary by model (common: 1:1, 4:3, 3:4, 16:9, 9:16, 21:9).

heightnumber

Desired height used to determine the aspect ratio and resolution tier. The actual output height is determined by the AI provider based on the closest matching supported aspect ratio. Supported ratios vary by model (common: 1:1, 4:3, 3:4, 16:9, 9:16, 21:9).

quality'480p' | '720p' | '1080p'

Video output quality/resolution.

audioboolean

Generate the video with audio or not. Some models (e.g. Grok Imagine Video / OVI) don't support the option to disable sound. With these models the result video will always be generated regardless to the provided input params.

lengthnumber

Desired video length in seconds. Mapped to the closest supported duration for the selected model.

model'urn:air:seedance:model:seedance:seedance-1.0-pro-text-to-video@1' | 'urn:air:seedance:model:seedance:seedance-1.0-pro-fast-text-to-video@1' | 'urn:air:seedance:model:seedance:seedance-1.5-pro-text-to-video@1' | 'urn:air:pika:model:pika:pika-text-to-video-v2.2@1' | 'urn:air:kling:model:kling:kling-v2-master-text-to-video@1' | 'urn:air:kling:model:kling:kling-v2-1-master-text-to-video@1' | 'urn:air:kling:model:kling:kling-v2-5-turbo-text-to-video@1' | 'urn:air:kling:model:kling:kling-v2-6-text-to-video@1' | 'urn:air:kling:model:kling:kling-v3-text-to-video@1' | 'urn:air:runway:model:runway:gen4.5-text-to-video@1' | 'urn:air:minimax:model:minimax:hailuo-02-pro-text-to-video@1' | 'urn:air:ltxv:model:ltxv:ltxv-2-text-to-video@1' | 'urn:air:ltxv:model:ltxv:ltxv-2-text-to-video-fast@1' | 'urn:air:ovi:model:ovi:ovi-text-to-video@1' | 'urn:air:xai:model:xai:grok-imagine-video-text-to-video@1' | 'urn:air:google:model:google:veo-2.0-text-to-video@1' | 'urn:air:google:model:google:veo-2.0-exp-text-to-video@1' | 'urn:air:google:model:google:veo-3.1-text-to-video@1' | 'urn:air:google:model:google:veo-3.1-fast-text-to-video@1' | 'urn:air:google:model:google:veo-3.1-preview-text-to-video@1' | 'urn:air:google:model:google:veo-3.1-fast-preview-text-to-video@1' | 'urn:air:wan:model:wan:wan-2.7-text-to-video@1' nullable

Optionally choose a specific AI model to use for this video generation. If not specified, a default model will be applied. Please note that the default model may change over time as Picsart continues to improve performance and accuracy. Any change to the default will be made only after thorough testing and validation to ensure it delivers better results. If you require consistent behavior or wish to evaluate different models on your own, we recommend explicitly setting this parameter.

Example request

{
  "width": 512,
  "height": 512
}

Response

Accepted.

status'success' | 'error' | 'processing'
inference_idstring nullable