Create a voiceover from a PDF, slides or images

A Voiceover is a narrated video built from documents and images you already have. Jellypod writes the narration, generates the audio, and times every visual you supplied to the finished script. It never redraws or replaces your material. Voiceovers share the Videos library in the sidebar with Magic Video and Shorts; filter by format to see only Voiceovers. To compare them with generated-visual videos, see choosing a video Product.

Your uploaded pages and images become the video's visuals. Choose Magic Video if you want Jellypod to generate new visuals from your content.

Create a Voiceover

1. Open Home and select Voiceover in the composer, or open Videos in the sidebar and choose Create Video > Voiceover.
2. Choose one to four characters and a narration language. Language starts from the language your selected voices share (narration language), and duration presets are in creation settings (the default is Auto).
3. Add up to 10 files: drag them onto the composer, click + to browse, or pick existing files from the Asset Library. Each file can be up to 100 MB, with at most 50 pages or images total across all files.
4. Optionally add direction for the audience, tone or key points.
5. Click Generate.

Generate is available as soon as you add a file. A file still uploading or processing continues in the background once generation starts. If a file fails to process, the Voiceover fails with that file's error.

With several characters, generation hands off between them at natural breaks and uses each at least once. Reassign individual paragraphs afterward in the editor.

Jellypod opens Videos and the new card shows its status. If generation fails, use Retry generation on the card: it reuses the files you already uploaded. See video generation failures for charges and causes. When the card is ready, open it to edit, render, share or download.

Files you upload become reusable sources in the Asset Library.

Supported files

For other presentation formats, export to PDF or PPTX.

Pages inside a document stay together in their original order. Jellypod arranges separate documents and images into an initial sequence.

Voiceover supports 50 document pages or images combined across all files. Each PDF page, presentation slide, or uploaded image counts toward that total. If you exceed it, remove pages or images before uploading. Splitting the same pages into several PDFs does not reduce the total. A PDF accepted by the Asset Library can still exceed Voiceover's limit.

For other file-processing errors, follow the file's error message. If Jellypod asks you to upload it again to prepare it for Voiceover, retry the upload. If it still fails, re-export the file as PDF, PNG or JPEG and upload that. If the Voiceover itself fails on the file, create a new one with the re-exported file.

Speaker notes

When you upload a .pptx, Jellypod reads each slide's speaker notes and uses them as context for that slide's narration, alongside the slide itself. Nothing needs turning on, and a deck without notes works as before.

Notes count as intent, not a script to read out. Jellypod narrates their substance in the Voiceover's voice and drops stage directions such as timing cues. Your direction outranks both the slides and the notes.

Jellypod reads speaker notes from PPTX files. A .pptx exported from Google Slides keeps them; a PDF export does not.

How visuals are framed

Jellypod shows a full page for context, then can focus on a readable graph, figure, table or section while the narration discusses it. A figure that is a separate image inside a document can appear without the surrounding page. Captions, labels and legends stay with the content they explain.

Framing never redraws your material. When a crop could lose content, Jellypod shows the whole visual over a matching solid color instead. After creation you can reframe a single visual in the editor.

Zoom, captions and transitions

• Zoom effect (settings menu beside the duration selector): Auto holds document and presentation pages still and gives photos a subtle zoom, including a figure pulled out of a page. Choose On to zoom every visual or Off to keep all still.
• Captions are off by default, since slides often show the text already. Turn them on in the same menu, or later from the Captions button in the editor.
• Transitions: generation adds a slide transition before every visual after the first. Edit, replace or remove any of them in the editor.

Web research

Narration can draw on web research for background beyond your files. To avoid it for one Voiceover, say so in your direction.

What a Voiceover leaves out

Your files are the visuals, so the creation flow has no visual style, brand kit, watermark, orientation or background music choice, and the result is always landscape. In the editor you can add a watermark and your own background music, but not apply a brand kit or switch the visual style.

Length and cost

Fixed durations range from 3 to 60 minutes, and Auto stays under three quarters of your plan's Voiceover cap. See plan limits for caps by plan.

A Voiceover costs 30 credits per minute of narration, and nothing for visuals because you supply them. Regenerating narration afterward is free. Auto length is priced once generation finishes; see understanding credits.

Related guides

• Creation settings: duration, language and remembered values.
• Video editor: edit narration, reframe, replace visuals.
• Video library and sharing: share, download and embed.
• Video generation failures: failed generations and charges.
• Choosing a video Product: Voiceover vs Magic Video and Shorts.