PlansFree; from $16/mo (annual billing) USD per person monthly; lower rates require annual billing

Free: Free; One media hour per month and 100 one-time AI credits; limited AI Speech trial.
Hobbyist: $16/month; Per person, billed annually; $24 month-to-month. Ten media hours and 400 AI credits per month.
Hobbyist (month-to-month): $24/month; Per person, month-to-month; ten media hours and 400 AI credits per month.
Creator: $24/month; Per person, billed annually; $35 month-to-month. Thirty media hours and 800 AI credits per month.
Business: $50/month; Per person, billed annually; $65 month-to-month. Forty media hours and 1,500 AI credits per month.

AI & setupAI includedFree plan includes 1 media hour

AI credits and media hours are separate monthly allowances; recording length and repeated AI processing draw on different limits.

Where it worksWeb, macOS, WindowsWeb and desktop editor

Edit audio and video in the same text-based workspace, with exports including MP3, WAV, and M4A.

Usage & termsMedia hours and AI credits are separateVoice rights depend on the selected voice

Paid plan prices vary between annual billing and month-to-month billing; the plan rows show both.

Free includes 100 one-time AI credits. Processing and media hours are budgeted separately.

Stock voice owners may remove their voices; Descript recommends its Recommended list for recurring projects.

Descript homepage with a maroon video-editing hero, an AI editing headline and a free-start button. — Descript homepage screenshot · captured 3 October 2026 Visit Descript (opens in a new tab)

Cut the recording through its words

Delete a digression from the transcript and its audio goes with it; move a sentence and move the recording too. Filler-word and retake tools identify possible cuts, while Underlord can assemble an edit from written instructions that remain reviewable.

Repair the sound and the spoken line

Studio Sound reduces background noise and echo. Regenerate can synthesize replacement speech to match surrounding audio, giving an already recorded passage a route to correction without rebuilding the entire edit.

Media hours and AI credits are separate allowances, so recording length and repeated AI processing affect the budget differently.

Direct more than one speaker

Paste a script and assign voices to selected passages for a reading or dialogue sample. Current speech models support tone tags for delivery such as whispering or laughter, with a documented language list that differs from transcription coverage.

The voice catalog includes ElevenLabs community voices that owners may remove. Descript recommends its Recommended list for recurring projects.

Export and plan allowances

Carry the edit into audio and promotion: Export MP3, WAV or M4A, including separate files at markers or line breaks and batch exports of compositions. The same workspace can prepare captions and social clips from the long recording, keeping promotional edits attached to their source.

Feature details & official sources

Purpose and capability information checked 3 October 2026.

Platform and pricing reference: 24 October 2026.

Recorded requirements and limits

Imported or recorded media consumes a media-hour allowance; AI tools use a separate credit allowance. Free access includes one media hour monthly and 100 one-time AI credits.
Select a voice compatible with the script language. Community voices can be withdrawn by their owners; stock-voice commercial scope was not resolved in the inspected documentation.
Voice cloning requires the speaker's authorization. Whole-book chapter production and retailer delivery specifications need a separate production workflow.

Information may be inaccurate. Confirm details on the official website. Report errors to [email protected].

Cut the recording through its words

Repair the sound and the spoken line

Direct more than one speaker

Export and plan allowances

Carry the edit into audio and promotion

Recorded requirements and limits