Generating an AI Voiceover
PublishFi can turn any text you write into a spoken voiceover using a library of AI voices. This is ideal for narration, explainer videos, how-to content, or any format where you want a polished spoken track without recording audio yourself. The voiceover panel lets you write your script, pick a voice, build multi-speaker compositions, and place the result on your timeline — all in one place.
Opening the Voiceover tab
Open the audio panel from the right-side panel of the editor. At the top of the panel, click the Voiceover tab. The panel opens to the Generate sub-tab by default. There is also a Library sub-tab for browsing and selecting voices — more on that below.
Selecting a voice
At the top of the Generate tab, the current active speaker is shown with their name and profile image. To change the voice, click the voice selector to browse the available options.
If you want to explore the full voice catalog, switch to the Library sub-tab. Browsing the library lets you preview voices and choose one that fits your content's tone — conversational, authoritative, warm, or energetic. Selecting a voice from the Library sets it as the active voice in the Generate tab.
Writing your script
The text area in the Generate tab is where you type or paste the text you want the AI to read. The placeholder reads: "Start typing here or paste any text you want to turn into lifelike speech." The textarea expands automatically as you write more.
Each block of text is associated with a specific speaker, shown above the input. This means you can build a multi-speaker composition within a single voiceover project — useful for dialogue, interview simulations, or narration that switches tone between sections.
Adding multiple speakers
To add a new text block with a different voice assignment:
- Hover over the separator at the bottom of a text block until the Add speaker button appears, then click it.
- Or press Shift + Enter anywhere in a text block to insert a new block immediately below without reaching for the mouse.
Each new block defaults to its own speaker slot, which you can change independently. There is no limit on how many blocks you can add to a single voiceover composition.
Generating and previewing
When your script is ready, click Generate. PublishFi sends the text and voice configuration to the AI engine and returns the audio. When generation is complete, a waveform player appears beneath the script.
The player includes:
- Play/Pause — start or stop playback
- Rewind backward — jump back in the audio to review a section
- Rewind forward — skip ahead
The player also shows a timestamp readout so you can see exactly where you are in the generated clip.
Adding the voiceover to the timeline
Click the + (Add to timeline) button in the player controls. The voiceover clip is placed at the current playhead position and appears in violet on the timeline, distinct from music clips (green), SFX (amber), and custom audio (blue).
Saving and revisiting compositions
Voiceover projects are saved automatically as you work. To return to a previous script, click the project selector at the top of the Generate tab. Saved projects are grouped by date and display a preview of the transcript text, the number of speakers, and how long ago the project was last updated.
To remove a project you no longer need, hover over it in the project list and click the delete button that appears.
Tip: Shorter scripts tend to produce cleaner results. For social video formats, aim for narration that covers one clear point per segment rather than long unbroken passages — it is easier to edit and easier for viewers to follow.
