Types into every app your show already runs on
Show notes, guest outreach, ad-read scripts, episode descriptions. Hold one key and say them the way you'd say them on mic. Clean, punctuated text lands at your cursor in any app, entirely on your Mac.
Press and hold the pill, then let go. That is the whole app.
Types into every app your show already runs on
Here's the irony nobody in podcasting says out loud: you built a career on being better at talking than everyone else, and then you run the entire business around the show at 40 words a minute, with your hands.
The conversation was the easy part. You were great in it, you always are. Now look at what stands between this exported file and an actual published episode. None of it is audio. All of it is typing.
Show notes. The most procrastinated task in podcasting, and it's not close.
TypedEpisode description. The one paragraph that decides whether a stranger presses play.
TypedGuest thank-you and outreach to the next three guests. Booking is sales.
TypedClip captions. Six clips, six captions, six little chores.
TypedNewsletter recap. The listeners who actually convert read this, not the app.
TypedTitles, sponsor follow-up, listener replies. The long tail that eats the evening.
TypedEvery item on that list is words you could say out loud, three times faster, in the voice that made the show worth listening to in the first place.
Timestamps below start the moment you hit export. Same episode, same show, same you. The only variable is whether the work around the mic goes through your hands or your voice.
Export starts rendering. You open a doc called "show notes ep 147" and stare at it.
Three sentences in. You check the render. You check your phone. The draft goes cold.
Guest thank-you typed, rewritten twice, still stiffer than you'd ever sound on air.
Clip captions. Six clips, six tabs. You start recycling the same caption with the nouns swapped.
The episode description gets one tired paragraph. It reads like a chore, because it was one.
The newsletter is a link and an apology. Outreach to next month's guests slips. Again.
An hour to record it. A day to type about it.
Export starts rendering. You hold ⌥Space and talk the show notes while the takes are still ringing in your ears.
Show notes done. The progress bar hasn't hit 20 percent.
Episode description, spoken like a cold open, because that's exactly what it is.
Guest thank-you, then two outreach emails for next month, in the same voice that hosts the show.
Six clip captions, one breath each. Newsletter recap right behind them.
Everything written. The export is still rendering. You go make coffee.
Show notes done before the export finishes. That's the whole pitch.
No new workflow, no template pack, no virtual assistant. Just the fact that you speak roughly three times faster than you type, applied to a job that is almost entirely made of words.
You already listen to the edit before you ship it. So narrate while you do: hold the key at each chapter, say the summary, drop the timestamp, keep listening. By the final chapter, the notes are done and you never sat down to "write show notes" at all.
Same trick before the interview: play the guest's other appearances at 1.5x and dictate your research notes and questions as they hit you. You walk in with pages of angles no other host has asked, and it never felt like prep.
This document was spoken in one pass. Nothing here was typed.
Māra Līcīte sold her company on a Tuesday and spent the next year wishing she'd read one clause more carefully. We get into the sale, the silence afterward, and what she'd never sign again.
Links: Māra's newsletter, the book she mentions at 51:20, and our sponsor's offer below.
The average person types around 40 words a minute and speaks around 150. Almost everything around your show, from the description to the sponsor pitch to the listener replies, is words. Run those two numbers across a season and the shape of your week changes.
Hunched at the desk after the edit, drafting captions with your thumbs, rewriting the outreach email a third time. Typing is the off-mic you.
The on-mic you, pointed at the show notes for once. Capitalization, punctuation, and cleanup happen while you talk.
An hour of typed admin becomes roughly twenty minutes of spoken admin. The rest goes back into booking better guests, or into having an evening.
Typical adult rates for conversational English, shown as an honest illustration, not a promise. Your rate will be personal. Yours, specifically, is probably higher. You talk for a living.
No app to open, no window to find. It lives in your menu bar and wakes when you hold the key, in whatever app your cursor happens to be in.
Hold ⌥Space anywhere: the YouTube description box, your hosting dashboard, Gmail, Notion, the newsletter editor.
Say the description like a cold open. Say the outreach email like you're closing a guest on a call. This is your home turf.
Release the key. Clean, punctuated text appears at your cursor, processed on your Mac. No spinner, no upload, no waiting.
Skim it, fix a word if you want, ship it. You're back to the edit, the next booking, or done for the day.
The episode has an editor. The clips have a template. Everything else, the outreach, the notes, the pitches, the replies, has been you and a keyboard at 11pm. Not anymore. Every feature below runs on your Apple Silicon, and the network is never part of the loop.
Booking is sales, and warm beats templated every time. Dictate outreach that references their actual work in your actual voice. Three personal emails in the time one templated blast used to take.
An ad read is a spoken thing. Draft it by speaking it, and the script lands already sounding like you on mic.
Clean for descriptions and outreach, Message for listener DMs, Voice for raw research notes. Switch from the menu bar.
Play the guest's other appearances at 1.5× and dictate questions as they hit you. Walk in with angles no other host has asked.
Teach it a guest's name once, before the episode, and every mention in your notes, description, and outreach is spelled right.
Narrate the chapters while you listen back. The most procrastinated task in podcasting, done before the export finishes.
Every dictation saved as text, searchable in seconds. Pull up last season's sponsor pitch and say the new one over it. The audio itself is never kept.
Six clip captions in six breaths. Three title options spoken while you pour coffee. The listener who wrote you four paragraphs gets four warm paragraphs back. If it takes a cursor, it takes your voice.
“You already know talking beats typing. You built a career on it. The only place you forgot was your own inbox.”The whole page, in one clip
“I sold sponsor slots over email for three years at forty words a minute and never once noticed the irony. Now I dictate the pitch, and it sounds like the show, and the show is the thing they're buying. Outreach goes out the same day instead of 'this weekend.'”
“I dictate research notes while I listen to a guest's other interviews at 1.5x. By the time we sit down I've got pages of questions nobody else has asked them. Guests notice. It's become the thing people say about the show.”
“True crime show notes are brutal: sources, timelines, corrections, sensitivity notes. I talk them through right after the edit while every detail is still loaded in my head. And nothing about an unreleased case ever touches a server.”
Think about what you'd dictate in a week: next month's episode topics, an unannounced guest, the rate you're about to counter a sponsor with. Cloud dictation ships that audio to someone else's servers. OpenWhisper does the entire job on your Mac, and the only network request it ever makes is downloading the model, once.
OpenWhisper runs Whisper through Core ML on the Apple Neural Engine, the chip your Mac already paid for. No servers to rent means nothing to slow you down while an export renders in the background, and no subscription line item to justify against the show's budget.
Transcription runs on dedicated ML silicon, not your CPU, so your editor and your export queue keep their headroom while you dictate.
A hardware-aware recommendation on first run. After that, dictation works in a basement studio, on tour, or on the flight to a live show.
A menu bar app that idles quietly and wakes when you hold the key. No Electron, no daemon farm competing with your session.
Fair questions. Here's the honest answer to each.
Whisper is genuinely strong with names in context, and the custom dictionary covers the rest. Teach it Māra Līcīte once, before the episode, and every mention across your notes, description, outreach, and newsletter is spelled right. Same for sponsors, your show's title, and your recurring segments.
Yes, and that's one of the best workflows on this page. It's push-to-talk: the mic only records while you're holding the key, there's no always-on listening. Put the playback on headphones and the mic hears only you, so you can narrate show notes or dictate questions mid-listen without pausing anything.
Whisper was trained on an enormous range of real speech, so most technical, medical, legal, and hobbyist vocabulary lands correctly straight away. For the terms that are genuinely yours, guest names, product names, your niche's inside language, the custom dictionary catches what the model doesn't, and it only has to be taught once.
That transcript is the episode. This is everything around it: the outreach that books next month's guests, the description that earns the play, the sponsor pitch, the captions, the newsletter, the replies. None of that has an audio file to transcribe, because right now it only exists as typing you're putting off. OpenWhisper is how it gets spoken instead.
Audio is processed entirely on your Mac and never stored, period. Transcribed text is saved locally so you can search your own history, and nothing about an episode, a guest, or a deal ever touches a server. No account, no telemetry, nothing phoning home.
You get 14 days with every feature, on a real publish week. No card, no account, no cancel-by-Friday trap. If show notes aren't getting done before the export finishes by your second episode, delete it and you've lost nothing.
Not a subscription you have to remember to cancel between seasons. One purchase, every feature unlocked, running quietly on the Mac you already record and edit on.
What does one mid-roll sponsor slot pay you? What did your best guest ever do for the show? Booking is sales, and the warm, specific outreach email is what lands the guest a template never would. If faster, warmer outreach ever books you one guest or closes one sponsor you'd have missed, this paid for itself many times over. And if it merely kills the typing, it's still cheaper than one more year of a $15-a-month cloud tool.
Try it free for 14 days on a real publish week: real show notes, real outreach, real captions. If it doesn't earn back your afternoons, don't buy it. Nothing to cancel, nothing to remember, nothing charged.
Cloud dictation rents you your own voice, month after month, forever, which is a strange deal to offer someone whose voice is the product. OpenWhisper is software the old way: you buy it once, it runs on your hardware, and it owes nothing to a server bill.
No seat licenses, no usage tiers, no price hike email between seasons. You already resent enough subscriptions on the show's card.
One-time purchase. Yours forever.
Your show exists because you're better out loud than almost anyone. The work around it, the notes, the outreach, the pitches, the replies, is all words, and right now you're producing those words at 40 a minute with the wrong instrument. The trial is free for two weeks, needs no card, and runs on a real publish week. If show notes aren't done before the export finishes rendering, delete it. If they are, it's $196.97 once, and the second shift is over for good.
Try every feature free for two weeks on real show notes, real outreach, real captions. No account to create, no card to enter, nothing phoning home while you decide. Keep it and it's $196.97, once. Delete it and it's gone.
Start free trial macOS 14 or later. Apple Silicon recommended.