Chop the speech. Keep the fire.

A Fire-Press tool Free: files up to 20 s

Drop the speech

Drop MP3, WAV or M4A files here, or click to browse. Several at once is fine, up to 20 seconds each.Add more speech: drop files here or click to browse.

  • First run downloads a speech model (about 75 MB) once. After that it is cached and works instantly.
  • Nothing is uploaded. It all happens in this tab.

Ready when you are. Drop a speech to start.

Click or drag across words. Double-click a word to hear it.

Waveform

speech selection or clip other clips

The waveform shows up here once a speech is loaded.

Scroll to zoom, shift+scroll or drag to pan, click to play from a point. Drag the pink handles to fix clipped syllables.

Import phrases

AI Rap Pack

Paste the AI’s reply here, or drop a .txt file on this panel. If it has a ### PHRASES line, only the lines under it are used. Otherwise every line counts as one phrase.

    Nothing imported yet. Each phrase will show every place it was found, with a checkbox and a preview button. Phrases the AI made up stay listed in red.

    Rhyme finder

    Clips rack 0

    Breathing room added around each word.

    The rack is empty.
    Click a few words, hit Enter, and watch it fill.

    Export

    Formats

    Project

    Save your transcripts and clips to pick up later. Audio is not stored in the file.

    How to use it

    1. Drop in a speech. MP3, WAV or M4A, up to 20 seconds each. Several files at once is fine.
    2. Wait for the words. They appear as they are found. The first run downloads a speech model once (about 75 MB), then it is cached and instant.
    3. Pick words. Click, or drag across a run. Shift+click extends. Double-click a word to hear it. Press Enter to make it a clip.
    4. Fix the edges. Drag the pink handles on the waveform, or nudge with the buttons. Loop it to hear the change as you go.
    5. Export. Download everything as a ZIP of WAV and MP3 files with a manifest.csv. Or head to Builder, type a lyric and let it stitch the takes together.

    Keyboard shortcuts

    SpacePlay or stop the selection or selected clip
    EnterAdd selection as a clip
    LLoop the selected clip
    ← →Shrink or extend the selection end by a word (Shift moves the start)
    [ ]Nudge clip start earlier or later by 10 ms
    ; 'Nudge clip end earlier or later by 10 ms
    DeleteRemove the selected clip
    /Jump to search
    EscStop playback, then clear the selection

    AI Rap Pack: the round trip

    1. Get the words out. Once a speech is transcribed, press Copy AI prompt in the Library (or Download AI prompt for ai-prompt.txt). It holds a ready-made prompt plus the full transcript. Download transcript saves just the plain text.
    2. Ask an AI. Paste the prompt into any AI chat. It writes a funny rap, then a ### PHRASES list of the exact phrases it used, one per line.
    3. Bring the phrases back. Paste the whole reply into Import phrases (or drop a .txt file on it) and press Find phrases. Each phrase shows every place it was said. Exact means word for word, close match means about one word off, and red not found means the AI made it up.
    4. Pick your takes. Preview with the play buttons, tick the ones you want (or Check all exact matches), then Add checked to Clips Rack. They land with the current padding, ready to fine-tune like any other clip.
    5. Download the sample pack. In Export, tick Name files from the phrase, type a ZIP name if you like, and press Download all as ZIP.

    Vox-Slice never sends anything anywhere. The transcript text only leaves this computer when you paste it into an AI yourself, and your audio never does.

    Free and Pro

    You are on Free: files up to 20 s. The length is measured from the audio itself, not the file size. A longer file is never cut short: Vox-Slice turns it away and tells you why, so trim it and drop it in again. Pro lifts the limit to 15 minutes per file. See Pro.

    About Vox-Slice

    Vox-Slice is a Fire-Press tool for chopping speech into word-sized samples, the first step towards a funny parody rap or any remix. It transcribes right in your browser, so your audio and transcripts stay on your device. Nothing is hosted, nothing is uploaded, and there is no account to make.