AI Voiceover on the Studio Timeline: Script-to-Voice Narration Guide (2026)
Write a script beside your shots and generate AI narration onto the Studio timeline: multi-language neural voices, one clip per sentence, translated language versions, SRT/VTT export.
AI voiceover in Beat2Cut is not a separate text-to-speech site you visit and then leave with an MP3. It is a script pane inside the Studio timeline: you write narration next to your shots, pick a voice, and the generated speech lands directly on a voice track of the same timeline that holds your picture, original audio, and music — as one clip per sentence, with the natural pauses between sentences preserved. All tools require a free account. Audio beat detection and beat cut lists are permanently free and unlimited; Free accounts get 150 one-time free credits at sign-up for scene/silence detection, narration, and exports (included with Pro).
Here is the workflow, what makes the sentence-clip model worth caring about, and where the costs actually sit.
Why "generate into the editor" beats "generate, download, import"
The traditional loop looks like this: write a script in a TTS website, generate, download an MP3, import it into your editor, discover the pacing does not match the picture, go back, regenerate, download again. Every iteration crosses two tools and one download folder.
Putting generation inside the timeline collapses that loop:
- The narration lands next to the shots it describes — you hear it against the picture immediately.
- Each sentence is its own clip, so matching narration to picture is dragging clips, not re-recording takes.
- The pauses you wrote into the script (sentence breaks, paragraph breathing room) arrive as real gaps between clips — content, not padding to be squeezed out.
That last point is easy to underrate. A single monolithic narration file forces you to cut into it blind whenever the picture changes; per-sentence clips mean the edit points already exist where speech naturally breaks.
The voices: multi-language, from named providers
Speech is synthesized by cutting-edge multi-language neural voice engines. You pick a voice and a language from the catalog (the list filters by language); you do not deal with a vendor console, and the catalog grows as voices are added.
Two transparency notes, because they answer the questions people actually ask:
- What goes online? Your script text and generation settings are sent to the speech provider; the generated audio is stored with your account so projects can reload it. Your video and music files do not go along — they stay in the browser for detection, editing, and export.
- Training? Your scripts and generated audio are not used to train our own speech models.
Step by step: from script to voiced timeline
1. Open your video
Open a video in the Studio timeline. It becomes a picture track plus its original audio as a separate clip you can mute, trim, or mix against the narration.
2. Let shot detection write your outline
Run scene detection (5 credits on Free, included with Pro) and the video becomes a shot-by-shot list — each detected shot is a row you can write narration beside. Skip this if you would rather write one continuous script; the shot rows are scaffolding, not a requirement.
3. Write, pick a voice, generate
Write or paste narration, choose language and voice, and generate. Generation requires sign-in and uses credits, quoted against your balance before running. The result lands on a voice track: one clip per sentence, real inter-sentence pauses intact.
4. Edit like clips, because they are clips
Drag a sentence to land exactly on a shot change. Trim a gap tighter. Split where you want a beat of silence. Adjust per-clip volume, fade a sentence in under music. Everything you can do to any timeline clip, you can do to a sentence of narration.
5. Need the same video in other languages?
The same script pane handles multi-language versions: the script is translated into an editable draft per language (English, Chinese, Spanish, and Japanese are in the current catalog), you review and correct each draft, then each language generates onto its own voice track — and export carries one selected language at a time. The full walkthrough is in the multi-language voiceover versions guide.
6. Export
- SRT / VTT subtitles — generated from the narration clips and their real timing. 5 credits a delivery on Free, included with Pro.
- WAV / MP3 mixdown — the timeline's audio as one file, for use in any editor. 5 credits a delivery on Free, included with Pro.
- Merged MP4 — the finished cut with picture quality kept and audio mixed: 10 credits on the Free plan, included with a Pro subscription.
What it costs, honestly
| Thing | Price |
|---|---|
| Beat detection and beat cut lists | Free & unlimited (after sign-in) |
| Scene detection, silence removal, cut-list delivery | 5 credits on Free, included with Pro |
| AI narration | Sign-in + credits, billed per generation and quoted first |
| Script translation for language versions | Sign-in + credits, by script length × target languages |
| SRT/VTT, WAV/MP3 | 5 credits a delivery on Free, included with Pro |
| Merged MP4 | 10 credits on Free; included with Pro |
Free accounts get 150 one-time free credits at sign-up — enough to hear a real script in a few voices and run your pre-editing passes. A Pro subscription provides 10,000 credits per 5-hour window and includes unlimited scene detection, silence removal, and MP4 exports.
Who this fits
- Tutorial and course creators — narrate against the exact frames being explained, then export subtitles from the same clips.
- Product and marketing teams — iterate script wording against the picture without a studio booking.
- Multilingual channels — one script, reviewed translations, per-language narration tracks: publishing Spanish and Japanese versions stops being a second production.
- Editors who finish elsewhere — generate and time narration here, export WAV plus SRT, and finish in Premiere, Resolve, or Final Cut with a cut list.
FAQ
Do I need an account to try the editor?
Yes, every feature needs a free account except beat detection, which runs before you sign in. On Free, scene detection, silence removal, and cut-list delivery cost 5 credits each from your 150 one-time free credits, and Pro includes them. Narration uses credits on every plan, because speech synthesis runs on third-party providers.
Can I fix one sentence without regenerating everything?
Each sentence is a separate clip, so timing and arrangement fixes are direct manipulation — drag, trim, split, delete. For wording changes, edit the script and regenerate; billing is per generation and quoted before it runs.
Which providers make the voices?
The platform integrates studio-quality multi-language neural voices. You simply pick voices and languages directly beside your timeline without managing external developer consoles.
Is my footage uploaded when I use narration?
No. Narration sends script text and settings to the speech provider. Source video and audio stay in the browser for detection, editing, and export encoding; generated narration audio is stored with your account.
Hear your script against your picture: open the Studio script pane . New accounts start with 150 one-time free credits.
Ready to Try Beat2Cut?
Create beat-synced cuts, auto-split your footage into scenes, and add AI narration — all on one timeline in your browser.