Batch Multimodal Video Generation
Batch generate videos by combining any mix of audio, video, images, and text through multimodal driving.
meitu batch video-multimodal-generate
Note: Batch processing is available from meitu-cli v2.1.1 onward. Please upgrade if you're on an older version.
Usage Examples
# Config file mode (config file required)
meitu batch video-multimodal-generate \
--config ./batch.video-multimodal-generate.yaml \
--output-dir ./outputs \
--jsonConfig File Example
version: 1
defaults:
outputDir: ./outputs
items:
- prompt: Generate a visualization video driven by music
audio: ./music/song.mp3
- prompt: Generate a new video referencing the style of a sample clip
referenceVideo: ./refs/style.mp4Parameter Reference
| Parameter | Required | Description |
|---|---|---|
--output-dir | Yes | Output directory |
--config | Yes | Path to YAML/JSON config file (config file is required) |
--concurrency | No | Type: number; Default: 3; Number of parallel tasks (recommended 1-2 for videos) |
--max-retries | No | Type: number; Default: 0; Number of retries on failure |
--skip-exist | No | Skip existing output files |
--dry-run | No | Preview the plan without executing |
--no-progress | No | Disable per-task progress logging |
--json | No | Output results in JSON format |
--json-output | No | Write results to a specified JSON file |
--skill-name | No | Server-side skill identifier |