Version generation workflow
How the New version sheet, credit estimate, and async AI generation work.
Creating a new document version is a two-step sheet: pick a type, then choose output language and confirm estimated credits.
Natural
Smooth spoken rewrite
Lecture
Dense, informative
Conversational
Dialogue style
Step 1: Choose version type
| Type | What it does |
|---|---|
| Original | Copy source text; translates if output language differs |
| Natural | AI rewrite for smooth, natural speech |
| Lecture | Dense, information-focused pacing |
| Conversational | Dialogue-style, podcast-like flow |
Each type shows a feature checklist (included vs not included) before you continue.
Step 2: Output language and credits
- Pick one of many languages
- Review estimated credits (loaded before you confirm, with a coins icon)
- Tap Create to add the version
When AI runs
| Scenario | Generation needed? |
|---|---|
| Original, same language as document | Often instant copy |
| Original, different language | Yes (translation) |
| Natural, Lecture, Conversational | Yes (always) |
Generation runs in the background after you create the version. The document header shows progress until status is completed or failed.
After creation
Speechdash switches to the new version when ready (or you select it from the header dropdown). Each version has separate playback progress.
Credits
Billed by token usage plus a 1 credit flat surcharge per generation. Failed generations should not charge incorrectly; contact support if the ledger looks wrong.