Create video generation job

post

https://api.macpaw.com/ai/v1/videos

Creates a new video generation job. The gateway routes to the provider selected by model
(provider/model format). Supported backends include
OpenAI Create video
and Google Gemini API (Veo).

The parameter set is provider-dependent: optional body fields, allowed seconds and size values,
response fields, and job status values may differ. Request and response shapes follow the OpenAI video API
where applicable as a common baseline.

Tool calling and prompt/context caching are supported when the selected provider and model expose
those capabilities (check model info for supports_function_calling and supports_prompt_caching).

Required body fields

model – Must be set using the provider/model format (e.g. openai/sora-2 or
google/veo-3.1-fast-generate-001). Do not omit the model.
prompt – Text that describes the video to generate (see the upstream provider reference for length limits).

Other parameters such as input_reference, seconds, and size are optional and may be ignored or rejected
depending on the selected provider and model.

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Body Params

Request body for video generation. Shape follows OpenAI Create
video where applicable;
Google Gemini API (Veo) is also supported via provider/model
in model. The set of accepted parameters varies by provider; model and prompt are required.

Request body for video generation. Shape follows OpenAI Create
video where applicable;
supported providers also include Google Gemini API (Veo).

The set of accepted parameters varies by provider and model — not every optional field applies to every
backend. model and prompt are mandatory for this gateway operation.

model

string

required

Video generation model id in provider/model format. Examples: openai/sora-2, openai/sora-2-pro
(OpenAI Create video);
google/veo-3.1-fast-generate-001 (Google Gemini API).
Required; must not be omitted.

prompt

string

required

length between 1 and 32000

Text prompt that describes the video to generate. Length and content constraints depend on the provider
and model.

input_reference

object

Optional reference that guides generation. Supported reference types and whether image_url or file_id is required depend on the provider. When supported, provide exactly one of image_url or file_id.

seconds

string

Clip duration in seconds when supported. Allowed values depend on the provider and model behind the gateway; different backends may accept different durations.

size

string

Output resolution (typically widthxheight) when supported. Allowed values depend on the provider and model; different backends may expose different resolutions.

Responses

200Video job created (queued or in progress). Response shape follows the OpenAI Video object where applicable; field presence and values may differ by provider (see examples).

400Bad Request - Invalid input parameters

401Unauthorized - Missing or invalid authentication token

402Insufficient credits

429Rate Limit Exceeded

500Internal Server Error