Documentation
Pandrator workflows
Choose the stages for your intended output.
- New users
- Operators
flowchart TB
A[Book or text] --> B[Extract and clean]
B --> C[Speech plan and voices]
C --> D[Generate and export audiobook]
E[Audio, video or subtitles] --> F[Transcribe or import captions]
F --> G[Review and optional translation]
G --> H[Export subtitles]
G --> I[Generate and export voiceover]Main project types
| Project | Input | Usual stages |
|---|---|---|
| Audiobook | Document or text | Clean, segment, prepare speech, generate, assemble, export |
| Subtitles | Audio, video or captions | Transcribe or import, correct, translate if needed, review, export |
| Voiceover | Audio, video or captions | Subtitle stages, speech generation, synchronization, mixing, export |
| Media editing | Video with a timed transcript | Review cuts, keep or remove ranges, render and check |
Work at your own pace
You can run a stage, inspect its result and continue later. Advanced settings allow you to include or skip optional stages. A source with good captions may not need transcription; a same-language subtitle project may not need translation.
For agent-driven work, inspect the project with pandrator_get_workflow and ask pandrator_recommend_next_steps when the next action is unclear. A workflow plan records selected inputs and settings before execution. Review any required confirmations and provider disclosures in that plan.
Small changes
Regenerate only the affected speech segments when practical. Export variants can reuse existing audio to change the caption style, audio mix or output format. A changed source, selected version or setting may require a new plan.