The product

How Reverbwell edits your podcast

Not a filter stack. Not a template. The AI was trained on podcast audio specifically: the patterns, the pacing, the noise profiles that matter to your format.

Raw file
MP3 / WAV / M4A
AI edit pass
8-12 min avg
Export ready
MP3 / WAV / RSS

From upload to episode, step by step

Upload

Drag MP3/WAV/M4A to the Reverbwell editor. Files up to 4 hours. Multi-track sessions: upload the stems separately, Reverbwell handles alignment.

AI edit pass

Runs in parallel: filler detection, noise classification, loudness analysis, silence scanning. On a 90-minute file, typical turnaround is 8-12 minutes.

Review

Every word is clickable. Delete a paragraph: the audio cuts too. No DAW knowledge needed. You're editing the script, not the waveform.

Export

MP3 at your chosen bitrate, WAV 24-bit, or direct RSS upload URL. Chapter markers export as ID3 tags or a Spotify Chapters JSON.

Training data

What the model was trained on

We trained on 10,000+ hours of podcast audio under strict data isolation. No customer recordings, no third-party scraping. Source audio was licensed from production houses and our own recording sessions, spanning interview, solo narrative, roundtable, and remote multitrack formats.

Abstract audio waveform visualization representing the training dataset

Training data composition

Interview format: 38%
Solo narrative: 31%
Remote multitrack: 22%
Roundtable: 9%

10,000+ hours total. No customer recordings included. All source audio licensed from production houses.

The actual numbers from our early-access program

~68%
Average edit time reduction
Based on self-reported edit hours from 40 early-access podcasters
4.2h
Median edit session before vs 22 min after
Same early-access cohort, 2-hour raw interview format
93%
Filler word detection recall
Internal benchmark on 500 hours of held-out podcast audio

Fits where you already publish

Buzzsprout Podbean Anchor / Spotify for Podcasters RSS.com Libsyn Direct download

Works with these platforms. Not an official partner of any platform listed above.

Common questions

MP3, WAV, M4A, and AIFF are all supported. Files up to 4 hours long. If you're uploading a multi-track session, export each stem separately and upload them together. Reverbwell handles track alignment.
Upload each host's recording as a separate stem. The AI aligns the tracks, reduces bleed between microphones, and normalizes each track independently before combining. This is the scenario the Studio plan was built for.
Your files are stored in your account until you delete them or your account closes. We do not use customer recordings to train the AI model. That training used only licensed and internally produced audio.
Yes. Before the AI runs, you can set filler word sensitivity (aggressive, moderate, or conservative), noise reduction intensity, and target LUFS. After the edit, the transcript editor lets you undo individual cuts before exporting.
Up to 4 hours per session on all plans. For episodes longer than 4 hours, split the recording into parts, process each, and combine in your DAW. We've never seen a podcast episode that long, but the limit exists for processing infrastructure reasons.