AutoVoiceover
AI narration and voiceover for videos · Windows
AutoVoiceover is a Windows app that narrates your videos. Pick a video and it studies the footage and the original audio, then writes a narration script in the style you choose. Edit the script line by line, pick a voice, and it dubs the narration and mixes it with the original sound into an MP4 with SRT subtitles. It suits travel, shop-visit and outdoor videos where the footage leads and a narration track is missing.
Hover to play the promo · narration, voice and subtitles generated by the app (in Chinese)
From footage to finished video
Pick a video
Drag it into the window. MP4, MOV, MKV, AVI, WebM and other common formats work.
Generate narration
Choose the narration density and style; the app studies the footage and audio and writes a script.
Edit line by line
The script is listed sentence by sentence. Double-click to edit; only edited lines are re-voiced.
Voice and export
Pick and preview a voice, then dub, mix with the original audio and export MP4 and SRT.
Features
Watches and listens
Speech in the original audio is recognised on your PC (SenseVoice, runs on the CPU, no GPU needed). The footage is analysed in segments by a cloud vision model, Qwen (Alibaba Cloud Bailian) by default.
Four density levels
From "High" down to "Minimal". The default, "Low", keeps narration to roughly 20–30% of the video so the footage and original sound can breathe.
24 preset styles
Grouped into travel & outdoors, lifestyle & shops, and explainers. Write and save your own style too, and add an opening hook that says what the video is about.
Two voice services
Use ElevenLabs or one of Alibaba Cloud Bailian's 88 system voices, and preview before you commit. Automatic tone adds emotion to each line based on the footage.
Mixed MP4 + SRT
Narration is mixed with the original audio, which ducks automatically while the narrator speaks. Export an MP4 and SRT subtitles, or burn the subtitles in and choose the export resolution and quality.
No paying twice
Each voiced line is cached, so re-voicing only generates the lines you changed. If a quota or key error stops a run, "Continue" picks up only the unfinished part.
Trial and activation: try it free for 2 days after installing, then activate Pro (lifetime or time-limited). AutoVoiceover has its own activation codes; codes for Formlo and our other apps don't work. Contact us for an activation code.
Requirements
- Windows 10 / 11 (64-bit), version 1607 or later
- No dedicated GPU needed; speech recognition runs on the CPU
- An internet connection: footage analysis, script writing and voicing run in the cloud
- Your own Alibaba Cloud Bailian API key; an ElevenLabs key too if you voice with ElevenLabs. Each provider bills you directly
- Installer about 251 MB; per-user install, no admin rights needed
Good to know
- Scripts are written by an AI model, so the wording changes every run and won't match a hand-polished script. Read through it before voicing.
- It occasionally reads out a shop sign in the footage; just edit that line.
- Cloud cost and time grow with video length. Try a short clip first for long videos.
Data and privacy
Speech in the original audio is recognised on your PC. For footage analysis, video segments are uploaded to the vision model you choose; script text goes to Alibaba Cloud Bailian; narration text goes to the voice service you choose. All of these use your own API keys and handle data and billing under each provider's terms.
API keys are stored encrypted on your PC. Activation sends only a hashed device identifier. Uninstalling never deletes the videos you exported.
Who it's for
Release history
This software includes FFmpeg under the LGPL. Corresponding source: GitHub Release.