narrator-ai-cli-video-generation

Install the CLI via pip, set your app_key with narrator-ai-cli config set, then follow the Fast Path: select a movie from resources movies, generate a script, upload it to get a file_id, and create a task with task create --mode hot_drama. Poll the task status until the video completes.

Adapted narration (hot_drama mode) uses existing movie clips and requires a movie_id, making it the fast path. Original narration (original_mix mode) generates new content from a topic and requires additional clip and audio generation steps before creating the video task.

Yes, the CLI provides standalone tts and voice clone commands. Text-to-speech converts provided text using a selected dubbing voice, while voice clone creates a custom voice from a sample audio file with a name and description.

Video generation is asynchronous and typically takes 5-15 minutes, so a processing status is normal. Use narrator-ai-cli task status TASK_ID --wait --interval 10 to auto-poll, and avoid checking more frequently than every 10-15 seconds to prevent rate limiting.

The tool requires Python 3.10 or higher. Its dependencies are typer, httpx[socks], httpx-sse, pyyaml, and rich, installed automatically when you run pip install from the GitHub repository.