Paraphrase
Edit recordings by deleting words, with a word-level transcript
About Paraphrase
Paraphrase is for podcasters, presenters, course creators and anyone who records spoken audio and wants to tidy it up. You record in the browser or upload audio or video, and the app produces a transcript where every word is linked to the exact slice of audio that produced it. Selecting a word selects its sound, so editing becomes a matter of choosing which words stay.
A one-button cleanup reads the whole transcript and proposes a cut that removes filled pauses, false starts, hedges and repeated points. You see what was removed and why, can restore any word with a click, and every action is reversible. Projects can hold several takes edited together as one transcript.
What sets it apart is that the output is always your own recording, cut down. Paraphrase never generates speech, so there is no AI voice, clone or overdub. Exports are available in common audio formats, and video projects can export MP4 with picture and sound cut at the same points.
Related tools
Paraphrase alternatives
Ardor
One AI agent that builds, deploys and runs your apps
Reimorph
Edit photos from a reference image using text prompts
ImageToSTL
Turn a photo into a 3D-printable model with AI
BlaBlaScribe
AI transcription, subtitles and dubbing from audio or video
Token360
One OpenAI-compatible API for text, image, audio and video
Image to Prompt
Turn any image into a detailed AI image prompt