Descript

Descript is an AI-powered audio and video editing platform founded in 2017. It takes an unconventional approach to media editing — rather than requiring users to manipulate timelines and waveforms directly, Descript generates a text transcript of recorded audio or video, and editing is performed by editing the transcript text. Users delete words from the transcript to cut them from the audio, rearrange sentences to restructure content, or use the AI to fill gaps and correct mistakes through Overdub, its voice synthesis feature. This makes Descript particularly accessible to podcast producers, YouTubers, and teams producing interview or documentary content who are more comfortable with text editing than traditional timeline-based tools. Descript also includes screen recording, video publishing, and collaboration features, positioning it as a broader content production tool beyond a simple editor.

How it works

Users import or record audio or video files directly into Descript. The platform generates a text transcript using AI speech recognition. Editors work directly in the transcript — selecting and deleting text removes corresponding audio from the recording, enabling fast content cuts without timeline navigation. Filler words (ums, ahs, silences) can be automatically removed. Overdub allows users to correct mistakes by typing replacement words that are synthesised in the speaker's cloned voice. Finished content can be exported as audio, video, or published directly to platforms.