Lyric Video Maker

Tap-to-sync and the export are free, with a small “Made with AIFlowMusic.com” mark in the corner. AI sync costs 1 credit per song and is refunded automatically if it fails; new accounts start with free credits to try it. Songs you made on AIFlowMusic get their timing without a credit once you have made any purchase. Videos from AI-synced lyrics and from your own AIFlowMusic songs have no mark.

Yes, with AI sync. Drop the song, paste the lyrics (or leave them out) and every line is timed for you, usually in 5 to 20 seconds; then you check the result and export. As an AI lyric video maker it costs 1 credit per song, refunded automatically if the sync fails. The free option is manual: you tap along as the song plays.

The song is sent to a speech recognition model (Whisper), which returns every word it hears with a start and end time. If you pasted the lyrics, they are matched against those words in order, so each of your words takes the time of the word it matches; words the model missed are placed between their neighbours. The song is deleted from our storage as soon as the timing comes back.

On our test songs — pop, rock, rap and a Spanish track — half of the AI-synced lines started within about 0.2 seconds of the voice, and most within half a second. Dense mixes and fast rap are harder: a few lines can be a second or more off, and a line nobody could hear is highlighted as a guess. Pasting the lyrics makes a big difference, because the AI then only has to find when your words are sung, not what they are.

Yes. Leave the lyrics box empty and press Transcribe: the AI writes down what it hears, with its timing, split into lines. Expect to correct some words — sung lyrics are much harder to recognise than speech — or paste the official lyrics for a cleaner result.

Speech recognition detects the song's language on its own; we tested English and Spanish songs. Languages written without spaces between words, such as Chinese or Japanese, don't line up word by word yet — use tap-to-sync for those.

Yes — the Karaoke style fills each word in your colour as it is sung. It uses the song as you upload it; for a sing-along without the lead vocal, remove the vocals first with the AI Vocal Remover and time the lyrics on the instrumental with tap-to-sync.

MP4 with H.264 video and AAC audio, in 1920×1080 (16:9), 1080×1920 (9:16) or 1080×1080 (1:1), at 30 or 60 frames per second. With a transparent background the export is a WebM with an alpha channel, because MP4 cannot carry transparency.

You can load recordings up to 30 minutes. AI sync takes songs up to 10 minutes and 25 MB. One export can be up to 10 minutes at 30 fps or 5 minutes at 60 fps, and a clip length and start point let you export just the part you want.

Only if you use AI sync: the song is uploaded for the speech recognition and deleted as soon as the timing comes back. Tap-to-sync, the preview and the export all run in your browser, and the video itself is never uploaded.

You do. The video is made on your device from your audio, lyrics and images, and we don't keep a copy. Make sure you have the rights to the song, lyrics and artwork you use.

On Windows and macOS, recent Chrome and Edge export MP4. Browsers without an MP4 encoder export WebM instead, and a few (older Safari, for example) can preview but not export — the page tells you which before you press Export. Long exports are faster and more reliable on a computer than on a phone.