How to get a Rev or Otter transcript into Premiere Pro

To get a Rev or Otter transcript into Premiere Pro, export it from the service with timestamps and speaker names included, convert that export into Premiere's transcript JSON format, and import it in the Text panel with Import Static Transcript. The conversion step exists because Premiere doesn't accept Rev's or Otter's export formats directly — and the number one reason this workflow fails is exporting the transcript without timestamps, which leaves nothing for Premiere to sync to.

This guide covers the vendor side: exactly what to export from Rev, Otter, and Descript, and which converter turns each export into something Premiere accepts. For the Premiere side — where the import lives, static vs corrected transcripts, and the classic error messages — see the companion guide on importing a transcript into Premiere Pro.

Already have your export? Convert it now: TXT to Premiere JSON for Rev and Otter text exports, DOC to Premiere JSON for Word-document transcripts, or SRT to Premiere JSON if you exported captions.

Why bother, when Premiere transcribes for free?

Fair question. Three reasons this workflow keeps coming up:

• You already paid for the good transcript. A human-reviewed Rev transcript has correct proper nouns, real punctuation, and clean speaker separation. Re-running Premiere's Speech to Text throws that away and re-introduces the errors you paid to remove.
• The transcript is the deliverable of record. In documentary and interview work, the approved transcript may have been through legal or the producer. The edit should be built on that text, not a second transcription that drifts from it.
• The recording is already transcribed in your meeting stack. Otter sat in the interview call; the transcript exists before the footage reaches the editor. Importing it beats redoing it.

Exporting from Rev

Rev delivers transcripts as documents (Word, PDF, plain text) and captions as SRT/VTT — which one you have determines the path.

1. Open the transcript in your Rev account and choose Export/Download.
2. Pick the text (.txt) or Word (.docx) option, and make sure timestamps and speaker names are enabled in the export options. Rev's timestamped text export puts each cue on a line with its timecode and speaker — exactly the structure a converter can rebuild timing from.
3. Convert it: the TXT to Premiere JSON converter reads Rev's timestamped text layout (speakers included) and writes the JSON Premiere imports. Got a .docx instead? The DOC to Premiere JSON converter handles Word documents directly.
4. In Premiere: Text panel > Transcript tab > ... > Import > Import Static Transcript, select the JSON.

If what you have from Rev is caption files (.srt/.vtt), skip the document route entirely — convert the SRT with the SRT to Premiere JSON converter. Captions already carry timing per cue, so the conversion distributes word timing across each caption's window.

Exporting from Otter

Otter's export lives on the conversation page, per Otter's own export documentation:

1. Open the conversation, select the Transcript tab, click the three-dot menu, and choose Export.
2. Export as TXT, and in the export options turn on speaker names and timestamps. This is the step people miss — Otter happily exports a clean-reading transcript with no timing at all, which is useless for syncing. (Some export options are gated to Otter's paid plans; the timestamped TXT is the one you want.)
3. Convert the TXT with the TXT to Premiere JSON converter. Otter's speaker-first cue layout ("Speaker Name 0:42 — text…") is detected automatically, and speaker turns survive into Premiere.
4. Import in Premiere via Import Static Transcript, same as above.

Otter can also export SRT directly. If you prefer that route — or you already have the SRT — convert it with SRT to Premiere JSON instead. The practical difference: Otter's SRT has already chopped the transcript into caption-sized cues, while the timestamped TXT preserves Otter's paragraph/speaker segmentation. For Text-Based Editing, either works; the TXT route usually reads more naturally in the transcript panel.

Exporting from Descript (and everything else)

Descript exports transcripts as Word documents, plain text, and SRT. The same decision tree applies: a timestamped text or Word export goes through the TXT or DOC converter; an SRT export goes through the SRT converter.

That decision tree is worth stating generally, because it covers services this guide doesn't name — Trint, Sonix, Happy Scribe, a station's in-house transcription, all of them:

• Timestamped text with speakers → TXT to Premiere JSON. Best segmentation, keeps speakers.
• Word document with timestamps → DOC to Premiere JSON.
• SRT/VTT captions → SRT to Premiere JSON (convert VTT to SRT first if needed).
• Text with no timestamps → stop. Re-export with timestamps on. Without timing there is nothing to import against — this is the wall, not a formatting quirk.

One habit worth building if you do this weekly: standardize the export settings once per service and write them into your prep checklist. "Otter: TXT, speakers on, timestamps on" is one line in a document, and it ends the every-project ritual of discovering the export was untimed after the editor has already asked where the transcript is.

A five-minute end-to-end example

Concretely, here's the whole trip for a one-hour Otter-transcribed interview:

1. In Otter, open the conversation → Transcript tab → three-dot menu → Export → TXT, with speaker names and timestamps on. You get something shaped like this:

Marta Reyes  0:04
So the first thing to understand about the project is that nobody expected it to work.

Interviewer  0:19
When did that change?

1. Drop that file on the TXT to Premiere JSON converter. It detects the speaker-first layout, reads each cue's timestamp, and writes a JSON where every word carries timing and each turn keeps its speaker. Download the JSON.
2. In Premiere, double-click the interview clip so it's in the Source Monitor, open Window > Text, pick the Transcript tab, then ... > Import > Import Static Transcript, and select the JSON.
3. The transcript appears in the panel, timed and speaker-labeled. Search for "nobody expected it to work," highlight the sentence, and Premiere assembles that exact clip range — Text-Based Editing, running on the transcript Otter made during the call.

Total time is genuinely about five minutes, most of it finding the export toggles. The batch version scales the same way: export all your interviews, convert them in one drop (a batch comes back as a ZIP, one JSON per transcript), and import per clip as you get to them.

What "good enough" timing means here

A converted transcript estimates word-level timing from the timestamps in your export — a cue every few seconds from Rev, per-caption windows from an SRT. That's plenty for what Text-Based Editing actually does: you select sentences, and Premiere cuts at the right frames because sentence boundaries carry real timestamps. What it is not is per-word scientific truth between those timestamps — words inside a cue are distributed across the cue's duration. In practice you will not notice while editing; if a specific edit point feels a frame or two off, nudge it on the timeline like any other cut.

One honest limitation worth knowing before you commit: the imported transcript is static. Premiere treats it as the transcript for that clip, and Text-Based Editing works fully — but it won't re-run speaker identification on it, and corrections flow through the corrected-transcript merge described in the import guide, not through re-transcription.

FAQ

Can Premiere Pro import an Otter or Rev transcript directly? No. Premiere's transcript import accepts its documented JSON format (and plain text only for correcting an existing transcript). Rev and Otter exports — TXT, DOCX, SRT — all need converting to that JSON first, which is what the converters above do.

Do speaker names survive the import? Yes, if they were in the export. Rev's timestamped text and Otter's speaker-first TXT both carry speakers, and the converter writes them into the JSON as Premiere speaker records. Export without speaker names and there's nothing to carry.

My export has no timestamps — can the converter guess the timing? No, and be suspicious of anything that claims otherwise. Timing has to come from somewhere real. Re-export from the service with timestamps enabled; every major service has the option, though some hide it in export settings.

Should I export SRT or timestamped text from Otter? Both work. The timestamped TXT keeps Otter's natural paragraph and speaker segmentation, which reads better in the transcript panel; the SRT is pre-chopped into caption-sized cues. If your end goal is captions anyway, going SRT end-to-end is the shorter path.

Does this enable Text-Based Editing on the clip? Yes — that's the point of importing as a static transcript. The words map to timecode, so highlighting text builds edits on the timeline, same as if Premiere had transcribed the clip itself.

What if I have the transcript as a PDF? Convert the PDF to text or Word first (any PDF tool does this), check that the timestamps survived, then run it through the TXT or DOC converter. PDFs with no timestamps hit the same wall as any other untimed text.