Free tool · In-browser · Stable Audio + MiniMax + Whisper + ffmpeg

AI video polish

Upload a screen recording or product demo. Add any combination of AI voiceover (Gemini writes the script, MiniMax speaks it), generated background music (Stable Audio, looped to fit), and English captions (Whisper transcribes, ffmpeg burns in or saves as sidecar .srt). Built to handle long videos — a 7-minute recording stays near-real-time when captions are off because we stream-copy the video stream and only touch audio.

Source video

Upload your video

MP4, MOV, WebM, MKV. Up to 1 GB. Any length.

Or paste a video URL

Direct video URL (S3, your CDN, Loom raw, Dropbox direct link). YouTube / Vimeo aren't supported — they block cross-origin fetch.

Replace audio with AI voiceoverGemini watches the video and writes a narration; MiniMax speaks it. Your original audio is replaced.

Add background musicStable Audio generates ~90s and ffmpeg loops it to cover the full video. Mixed under voice if voice is on.

Mood

Music volume (-14 dB)

SubtleProminent

Add AI intro clipHappy Horse restyles the first 6s of your footage into a short branded opener, prepended to the video. One generation (~$4, 720p).

Viral YouTube modeSwitches the voiceover prompt to a hook-retention-CTA structure, and writes a YouTube title, description, and tags after the polish so you can paste into Studio.

Captions

OffNo captions.Sidecar .srt fileFast. Download a separate .srt the user can toggle on in YouTube / their player.Burn into videoCaptions baked into pixels. Re-encodes the video — slow for long clips.

Polish a video

Upload a screen recording or product demo. Pick what to add — AI voiceover, background music, captions — and we'll mix it all into one downloadable file. Designed for long recordings: a 7-minute video processes in seconds when captions are off, in single-digit minutes when on.