For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Python class

OutputVideoContent

OutputVideoContent

class max.pipelines.request.OutputVideoContent(*, type='output_video', video_url=None, video_data=None, format=None, frames_per_second=None, num_frames=None, frames=None)

source

Bases: BaseModel

Video content generated by the model in output messages.

Parameters:

• type (Literal['output_video'])
• video_url (str | None)
• video_data (str | None)
• format (str | None)
• frames_per_second (int | None)
• num_frames (int | None)
• frames (Annotated[ndarray[tuple[Any, ...], dtype[uint8]] | None, SkipJsonSchema()])

format

format: str | None

source

frames

frames: SkipJsonSchema[npt.NDArray[np.uint8] | None]

source

frames_per_second

frames_per_second: int | None

source

from_numpy_frames()

classmethod from_numpy_frames(frames, *, frames_per_second=None, format=None)

source

Create internal video content from raw frame data.

Parameters:

• frames (ndarray[tuple[Any, ...], dtype[uint8]]) – A uint8 array with shape [T, H, W, C] containing raw video frames.
• frames_per_second (int | None) – Optional frame rate metadata to attach.
• format (str | None) – Optional output format metadata, typically "mp4".

Returns:

An OutputVideoContent instance carrying raw frames for later route-level encoding.

Return type:

OutputVideoContent

model_config

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True, 'frozen': True}

source

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

num_frames

num_frames: int | None

source

type

type: Literal['output_video']

source

video_data

video_data: str | None

source

video_url

video_url: str | None

source