lexigram-multimedia-video

Video generation and processing for the Lexigram Framework — local and API-based backends (local-http, runway, openai, wan22, cogvideox, svd, comfyui) plus an ffmpeg-backed VideoProcessor.

Overview

lexigram-multimedia-video generates video from a text prompt and edits existing clips. The default backend calls a local HTTP reference server (http://localhost:5004) so the package works out of the box with no API keys; hosted backends (Runway, OpenAI), ComfyUI workflows, and local reference servers (Wan2.2, CogVideoX, SVD) are selectable via config. A built-in VideoProcessor provides ffmpeg-based editing (trim, concat, overlay, transcode, subtitles, GIF, and more).

Full documentation: docs.lexigram.dev

Install

uv add lexigram-multimedia-video
# Optional extras
uv add "lexigram-multimedia-video[runway]"            # Runway API
uv add "lexigram-multimedia-video[openai]"            # OpenAI Sora API
uv add "lexigram-multimedia-video[cogvideox-server]"  # local CogVideoX server (torch)
uv add "lexigram-multimedia-video[svd-server]"        # local SVD server (torch)

Quick Start

from lexigram import Application
from lexigram.di.module import Module, module
from lexigram.multimedia.video import VideoModule
from lexigram.contracts.multimedia import VideoProvider, VideoRequest


@module(imports=[VideoModule.configure()])
class AppModule(Module):
    pass


async def main() -> None:
    async with Application.boot(modules=[AppModule]) as app:
        video = await app.container.resolve(VideoProvider)
        result = await video.generate(
            VideoRequest(prompt="a drone flying over mountains")
        )
        if result.is_ok():
            asset = result.unwrap()  # MediaAsset — video bytes or URI


if __name__ == "__main__":
    import asyncio

    asyncio.run(main())

Configuration

Zero-config usage: Call VideoModule.configure() with no arguments to use the local-http backend at http://localhost:5004.

Option 1 — YAML file

# application.yaml
multimedia:
  video:
    backend: "comfyui"
    comfyui_base_url: "http://localhost:8188"
    processing:
      max_concurrent_jobs: 4

Option 2 — Profiles + Environment Variables

export LEX_PROFILE=production
export LEX_MULTIMEDIA__VIDEO__BACKEND=runway
export LEX_MULTIMEDIA__VIDEO__PROCESSING__MAX_CONCURRENT_JOBS=4

Option 3 — Python

from lexigram.multimedia.video import VideoModule
from lexigram.multimedia.video.config import VideoConfig, VideoProcessingConfig

VideoModule.configure(
    config=VideoConfig(
        backend="comfyui",
        processing=VideoProcessingConfig(max_concurrent_jobs=4),
    )
)

Config reference

Module Factory Methods

Key Features

• Seven generation backends — local-http, runway, openai, wan22, cogvideox, svd, comfyui
• VideoProcessor — ffmpeg-backed editing: trim, concat, overlays, composition, subtitles, audio mux, thumbnails, GIF, transcode, speed, crop, color filters, raw filter graphs
• Reference servers — lexigram-video-*-serve console scripts run each local model server
• Secret-managed API keys — provider keys resolved by name through the secrets backend
• Result-based — generate() -> Result[MediaAsset, MultimediaError]; errors are domain values, not exceptions

Testing

from lexigram import Application
from lexigram.multimedia.video import VideoModule


async def test_boot():
    async with Application.boot(modules=[VideoModule.stub()]) as app:
        assert app.container is not None

Key Source Files