Descript is a text-based audio and video editor from Descript, Inc. that lets you edit video by editing a transcript, with AI transcription in 25 languages, AI voice cloning, and dubbing, plus an AI co-editor called Underlord for automating edit tasks. It offers a genuine free plan (60 transcription minutes/month) and paid plans starting at $16/person/month billed annually.
Who builds it
Descript is built as a "type to edit" video and podcast editor: instead of scrubbing a timeline, you cut, rearrange, and clean up footage by editing the auto-generated transcript, and Underlord — Descript's built-in AI — can execute editing instructions directly, such as removing filler words, following a rough script, or generating rough cuts. Descript also exposes an API and an MCP server, so editing can be triggered from code or from inside another AI assistant.
Core features
- Transcript-based editing — cut video or audio by deleting or rearranging text, with changes reflected in the underlying media automatically.
- Underlord, the AI co-editor — automates edits like filler-word removal, following a script, and building rough cuts from raw footage.
- AI voice cloning and dubbing — clone a voice or translate/dub a project into another language.
- Automatic transcription in 25 languages, the basis for the text-based editing workflow.
- API + MCP — edit video programmatically or from inside an AI assistant, not just through the app.
- 4K, watermark-free export as a paid-tier capability, per Descript's own pricing page.
Pricing
Descript is freemium: the Free plan includes 60 transcription minutes per month at $0. The next tier, Hobbyist, starts at $16/person/month billed annually ($24 billed monthly), and unlocks 4K, watermark-free export among other features. Higher tiers add more transcription minutes, seats, and collaboration features; Descript's pricing page does not spell out a resolution cap for the Free plan specifically, since it's scoped by transcription minutes rather than export settings.
Who it's for
Descript fits podcasters, YouTubers, and video/audio teams whose primary editing task is cutting spoken content — removing filler words, trimming interviews, or repurposing long recordings — where transcript-based editing is faster than timeline scrubbing. The API/MCP access also makes it a fit for teams that want to trigger edits programmatically from their own tools. It's a narrower fit for people who need motion graphics, generative AI-video creation, or heavy visual effects work, where a general-purpose or generative video tool is a better match.