Free tool · Gemini 2.5 Flash + MiniMax + ffmpeg.wasm
AI voiceover for your demo video
Upload a raw Shopify app demo. The tool samples keyframes, transcribes any existing audio, generates a voiceover script with Gemini, speaks it in a voice you pick, and merges the new audio onto your video into one downloadable MP4.
1
Upload video + analyze
Drop in your Shopify app demo. We extract 6 keyframes in the browser, transcribe any existing audio with Whisper, and ask Gemini 2.5 Flash (multimodal) to write a 60-second voiceover grounded in what's actually on screen.
2
Review the script
Tweak the wording, pick a voice, and generate the voiceover audio.
Voiceover script—
3
Merge + download
ffmpeg.wasm swaps the audio track on your video with the new voiceover (runs in-browser, ~25 MB core downloads once).
Pipeline
- Keyframes — 6 evenly-spaced frames captured in-browser from your video.
- Whisper — transcribes any existing audio so Gemini can ground the script in real narration, not guessed-at UI.
- Gemini 2.5 Flash — multimodal model reads the frames + transcript and writes a ~60-second voiceover with timed segments and rules against superlatives, stats, pricing claims, and emojis.
- MiniMax Speech 02 HD— synthesises the script in a preset voice.
- ffmpeg.wasm — swaps the audio track on your original video, encodes H.264 + AAC, and hands you the final MP4. Everything stays in your browser.
Pair with
Record your demo with the screen recorder, then drop it straight into this tool.