what descript is
descript is a podcast and video editing studio. its central insight, and a genuinely original one, is that editing audio and video is fastest when you edit the transcript instead. delete a sentence in the transcript and the audio cuts to match. a remove-filler-words pass removes them in both places at once. AI voice cloning lets you patch a misspoken phrase without re-recording. screen recording, multitrack editing, publishing, hosting — descript is a vertically integrated content pipeline.
descript's transcription is good and getting better. but the transcript isn't the deliverable for descript users. the edited podcast is. the published video is. the transcript is the editing surface — a means, not an end.
what we are
audiohighlight is a transcript workspace. the transcript is the deliverable. journalists publish from it. researchers code it in NVivo. paralegals file deposition format from it. podcasters use it for show notes and citations. therapists paste excerpts into clinical records. conversation analysts publish jefferson-notation transcripts as primary research data.
what we optimize for is the gap between what the model delivers and what the user needs to use — speaker labels you fix in bulk, every word linked to its second of audio, exports shaped to the downstream tool, custom vocabulary the model learns once and remembers across files.
which one fits you
descript fits if:
- you're producing podcasts or videos and the deliverable is the edited content
- you want a unified pipeline: record, edit, mix, publish, all in one tool
- you want overdub / voice-cloning to patch misspoken phrases
- you're collaborating with a small content team that all works inside descript
- screen recording and multitrack editing are part of your workflow
audiohighlight fits if:
- the transcript is the deliverable, not the edit surface
- you need a transcript shaped for a specific downstream tool (NVivo, ATLAS.ti, MAXQDA, deposition format, jefferson notation)
- you need on-device transcription for HIPAA, privilege, or source-protection reasons (descript has no equivalent)
- you're a journalist, researcher, paralegal, therapist, or conversation analyst — not a content creator
- you're paying out of pocket and a $30/month subscription doesn't fit your usage shape
where descript is genuinely better
- video and audio editing. we don't do this at all. if you're editing a podcast or a youtube video, descript is the right tool, full stop.
- filler-word removal at scale. descript's automatic-filler-word removal is a real feature with real downstream value. we don't have it; if it's the workflow, descript wins.
- overdub / voice cloning. patching misspoken phrases without re-recording is unique to descript. we won't compete on this — it's not the problem we care about.
- publishing and hosting integration. descript publishes podcasts and videos directly. we hand you a transcript export and stop.
where we're different
- format-specific exports. descript exports a transcript in standard formats (.docx, .srt, .vtt, .txt). we export deposition format, NVivo CSV, jefferson notation, MAXQDA-shaped CSV — the formats that don't work with descript's content-creation use case but matter immensely for ours.
- private mode (on-device). descript uploads audio to its servers. for any audio that can't go there — therapy, privileged interviews, source material — that's a deal-breaker. our private mode runs the model in your browser with no upload.
- pay-per-file pricing. descript is subscription-based ($16/mo creator, $30/mo pro, $50/mo enterprise) with limits per tier. we're $0.20 per minute, pay per file, no subscription, all features included. for occasional / individual use, that's a much smaller commitment.
- transcription as deliverable, not surface. our editor optimizes for citation, verification, and handoff to a downstream non-descript tool. descript's optimizes for editing-the-content. when the transcript is the deliverable, the optimization runs in different directions.
can you use both?
yes. several podcast producers use descript to edit the show and audiohighlight to produce the show notes, episode transcripts for the website, and deposition-style transcripts for guests' legal/PR review. the two tools don't conflict — they answer different questions about the same audio file.
switching workflows
descript exports as .docx, .srt, .vtt, .txt. drop those into our editor and we re-format without re-transcribing. so a show transcript edited in descript can become an NVivo-shaped or deposition-format export here without re-paying for transcription. the editor handles cross-tool handoffs as a first-class case.