AI voice-over for your videos

Write or draft the script, pick a voice, and the narration lands on your timeline.

How it works

Choose what to narrate Select clips, set an in and out range, or start at the playhead in the video editor.
Write the script Type it yourself or let the AI draft one from what is in the picture, then see how long it will run.
Pick a voice and place it Choose the voice and quality; the narration lands on an audio track and the original sound is turned down under it.

Twelve voices Coral, Nova, Sage, Alloy, Ash and Onyx in Standard; Sarah, Laura, Aria, George, Brian and Daniel in Studio quality.
Script from the picture The AI writes a first script from the clips you selected, at no cost.
Ducking built in Under the narration the clip's own sound drops to 30 percent, so the voice stays clear.
Direct the delivery Standard voices take a short direction, for example calm and slow, or upbeat like an ad.
Any language you write in Both qualities read the language of your script, no setting needed.
Length before you pay An estimate of the spoken length shows whether the script fits the shot.

Standard: OpenAI gpt-4o-mini-tts Six voices with a delivery direction, 0.04 credits per 1,000 characters.
Studio: ElevenLabs Multilingual v2 Six natural voices for finished productions, 0.20 credits per 1,000 characters.

Explainers and tutorials A clear voice that walks the viewer through each step.
Short-form video Reels and Shorts with a hook spoken over the first second.
Product videos Narration over product shots and AI clips, in your brand's tone.

0.04 credits per 1,000 characters in Standard quality and 0.20 in Studio quality, at least 0.01 credits. A 60-second script has about 900 characters.

Up to 4,000 characters per narration, about four minutes of speech.

Onto an empty audio track of your project, or a new narration track, lined up with the clips you chose.

Through the ContentMaschine API, yes. In the app the narration lives in the video editor.