AI podcast production
Turn a recording into a publish-ready podcast package.
Upload your recording, audio or video. Podra masters the loudness (-16 LUFS stereo, -19 mono), trims the silence, transcribes it with speakers, suggests the filler and pause cuts for you to accept, and drafts chapters, clips, show notes and titles. You edit, then export the files or publish to your own RSS feed.
Your first 30 minutes of audio are free. Plans from $19 a month, with a 7-day free trial.
For independent podcasters and small studios making interview, solo and video shows.
0:22 · Maya (host)So, uh, why bikes? You were an accountant before this, right? filler · Cut / Keep
0:27 · Theo (guest)Right, for eight years. … I thought, um, people keep knocking on my door. filler · Cut / Keep
- 0:00 Meet Theo and his garage bike shop
- 0:22 From accountant to bike mechanic
- 0:45 Counting before committing
- 1:08 The pricing mistake, and the booking fix
Sample data
The recording is done. The episode isn't.
After every recording there are hours of chores before anyone can listen: levelling the volume so it doesn't jump between voices and apps, cutting the ums and the dead air, writing show notes and chapters, finding a clip for social, and getting it all into a feed.
None of it is creative, and all of it decides whether the episode goes out this week.
Podra does the first pass of all of it, and leaves every decision to you.
How it works
From recording to package in four steps
Upload the recording
MP3, WAV, M4A, FLAC or a video (MP4, MOV), or one track per speaker (up to 4) for exact speaker labels. Podra checks the length against your plan before it uploads.
Review the suggested cuts
The transcript is the timeline. Fillers and long pauses are highlighted with Cut or Keep; select any words to cut them; play with your edits before exporting.
Edit the drafts
Chapters with times, 3 to 5 clips, show notes, title options and a description are drafted from the transcript. Edit, reject or regenerate any of them.
Export or publish
Download the package (MP3 or WAV with ID3 chapters, transcripts, notes, clips), or add the episode to your show's RSS feed when you choose to.
What you get
One upload, the whole package
Everything below comes out of a single recording, and every piece is yours to edit, reject or regenerate.
Audio
- Loudness mastered to podcast targets
- Silence trimmed
- Multitrack
- Video podcasts
Transcript and text-based editing
- Timestamps and speakers
- Fillers and long pauses, suggested
- Cut by deleting words
- Fix the words
Drafts you edit
- Chapters
- 3 to 5 clips
- Show notes, titles and a description
- What was used, and how
Export
- The episode
- Chapters
- Transcript
- Show notes
- Clips
- One ZIP
Who it's for
Made for the show you make
Accuracy, data and control
Built so nothing goes out you didn't choose
- Suggestions, not edits.
- Every filler and pause cut waits for your Cut or Keep. The original master is kept, so any cut can be undone.
- Drafts say where they came from.
- Notes, chapters and clips are labeled as drafted by AI or by rules, and only use what's said in the episode.
- Publishing is yours.
- The feed is private until you publish it, each episode joins it only when you add it, and Podra never submits to directories or posts on your behalf.
- Limits you can see.
- Minutes left, the longest file accepted and when minutes come back are shown before you upload. No overage charges: uploads simply wait.
Transcription is by OpenAI's speech-to-text and drafting by Anthropic's Claude; neither may train on your audio or transcripts. Transcripts are encrypted at rest. Data policy.
Pricing
Priced by audio hours, every feature on every plan
Plans differ by hours of audio processed a month, the longest file, shows and seats. An episode's length counts once, when Podra processes it (rounded up to the minute; multitrack episodes count the length of the mix, not each track). Re-editing, re-exporting and regenerating notes don't count again. During the 7-day trial, up to 120 minutes can be processed.
Creator
$19/month
For one show, about one episode a week.
- 4 hours of audio a month
- Files up to 2 hours long
- 1 show with RSS feeds
- 1 seat
Pro
$49/month
For a weekly long-form show, or a few shows.
- 15 hours of audio a month
- Files up to 3 hours long
- 3 shows with RSS feeds
- 2 seats
Studio
$99/month
For a small studio or network with a team.
- 40 hours of audio a month
- Files up to 4 hours long
- 10 shows with RSS feeds
- 5 seats
Prices in US dollars, billed monthly; cancel any time. When a period's hours are used, uploads wait for the next period or a bigger plan; nothing is charged extra. Every plan starts with a 7-day free trial (up to 120 minutes of audio during the trial). Before a plan, your first 30 minutes are free.
Questions
What does Podra do?
Podra turns a podcast recording into a publish-ready package. It masters the loudness to -16 LUFS (stereo) or -19 LUFS (mono), trims silence, transcribes with timestamps and speakers, suggests filler and pause cuts you accept or reject, and drafts chapters, clips, show notes and titles you can edit. Then you export the files or publish to an RSS feed Podra hosts.
Can I try it without an account?
Yes. The sample page processes a bundled sample with Podra's real mastering and cuts, with your choice of loudness target and cuts, and shows the transcript, chapters, clip and notes it produces. The sample's audio is synthetic, and the page says so. Uploading your own recording needs a free account.
How much does it cost?
Creator is $19 a month for 4 hours of audio, Pro $49 for 15 hours, and Studio $99 for 40 hours with 5 seats. Every plan starts with a 7-day free trial, and your first 30 minutes are free before you choose.
What counts against my hours?
An episode's length counts once, when Podra processes it (rounded up to the minute; multitrack episodes count the length of the mix, not each track). Re-editing, re-exporting and regenerating notes don't count again. During the 7-day trial, up to 120 minutes can be processed. Re-editing and re-exporting never count again.
Does it cut the ums automatically?
No, on purpose. Podra finds fillers (um, uh, erm) and pauses over 1.6 seconds and marks them in the transcript with Cut and Keep. You can accept them one at a time or all at once, and undo any cut, because the original master is kept.
How accurate is the transcript?
It's made by OpenAI's speech-to-text, which is accurate on clear speech and weaker on heavy crosstalk, strong accents or noisy rooms. Names are often misheard; you can correct any words in the editor. Speaker labels are exact when you upload a track per speaker, and automatic (worth checking) from a single file.
Does Podra publish to Spotify or Apple Podcasts for me?
No. Podra hosts an RSS feed for each show and tells you what directories need. You submit the feed address to Apple Podcasts, Spotify and others yourself, once, and new episodes you add to the feed then appear there. Podra never posts anywhere on your behalf.
Which file types can I upload?
Audio: MP3, WAV, M4A, AAC, FLAC, OGG and Opus. Video: MP4, MOV and WebM. Up to 4 GB a file. Multitrack: 2 to 4 audio tracks, one per speaker. The longest recording depends on your plan: 2 hours on Creator, 3 on Pro, 4 on Studio.
Which AI does it use, and is my audio used for training?
Transcription uses OpenAI's speech-to-text models; chapters, clips, notes and titles are drafted with Anthropic's Claude from the transcript. Neither provider may train on your content. Transcripts are encrypted at rest, and you can export or delete everything.
Can it make video clips?
Yes. Each clip you accept is exported as an MP3, captions (SRT) and a captioned vertical 9:16 video, cut from the picture for video podcasts or over your show artwork with a waveform for audio. Fillers you cut are cut from the clips too.
Who do I contact with questions?
Email clio@new-odyssey.com. A person reads every message.
Your next episode, ready to publish
See it on the sample first, no account. Then upload your own: the first 30 minutes are free.