Talk it. Ship it.
Your best ideas live in voice memos that nobody reads. Cassette turns a 30-second rant into a tweet thread, a LinkedIn post, a blog draft, and a newsletter section — in your voice, not a template.
Voice memos are a graveyard.
You record the gold. You never publish it. It sits in iCloud rotting next to a 4-year-old grocery list.
Recording is easy.
You think while you walk. You riff in the car. You debrief after a call. Hundreds of memos. Ideas you'd kill to publish.
Editing is the wall.
Otter and Descript hand you a wall of text. Now you stare at 1,200 words and another two hours of work to make it shippable. So you don't.
The voice gets lost.
The few times you do publish, ChatGPT smooths every interesting edge off your thinking. It reads like a LinkedIn boilerplate. So you stop.
Hit record. Get four polished assets back.
Cassette listens to how you actually talk and writes back in that voice — with your vocabulary, your rhythm, your specific examples preserved.
Record anywhere.
Open Cassette on your phone or laptop. Hit the red button. Talk for 30 seconds to 5 minutes. Stream of consciousness is fine — Cassette likes it.
Live transcription.
Browser-native speech recognition transcribes as you talk. No upload. No "processing… 78%". You see the words appear in real time.
Four assets, one click.
Stop recording. Cassette generates a tweet thread, LinkedIn post, blog draft, and newsletter section — each preserving your voice and specific points.
Copy. Paste. Ship.
One-tap copy on each asset. Paste into X, LinkedIn, Substack, Beehiiv. Optional "Made with Cassette" footer is a toggle, not a tax.
Hook → payoff, ready to ship
2/ Every founder I know has 200+ voice memos with their best thinking. None of it is published. None of it ever will be — because editing kills it.
3/ The fix isn't another writing tool. The fix is a tool that respects the original voice...
Founder voice, scroll-stopping opener
Almost none of them have ever become anything. They're not bad ideas — they're some of the best thinking I've done all year. Recorded on a walk between meetings, after a customer call, the morning after a bad investor pitch...
Structured, with H2s and a takeaway
## What's actually happening
Most knowledge workers — especially founders, operators, and creators — are recording the best ideas they'll have all year. And then doing nothing with them.
## Why traditional tools don't fix it...
Scannable, ready for Substack
You have a content problem. Not the kind you think.
The problem isn't that you don't have ideas. You have hundreds. They're sitting in your voice memos, recorded on walks, after calls, in airports...
Built for people who'd rather talk than type.
Founders building in public
Daily threads keep the audience compounding. Cassette lets you keep momentum without burning two hours of writing time.
Operators with sharp opinions
You debrief after every call with takes nobody else has. Capture them between meetings, ship them at lunch.
Creators who hate the keyboard
Some people think out loud. Some people think on the keyboard. Cassette is for the first kind.
Execs who travel constantly
Walk through SFO, dictate the week's POV, land with four drafts. Pin it in a Notion page later if you want.
I had 247 voice memos on my phone with the best ideas I'd had all year. Not one of them ever became anything. I built Cassette so the best things I think out loud actually leave my head.
It's also a voice infrastructure bet. The same engine that turns memos into threads is going to turn calls into briefs, pitches into decks, and standups into changelogs. Cassette is step one.
Free to start. Real plans for real volume.
If you publish more than once a week, Creator pays for itself the first day.
Free
Five recordings a month. Every output unlocked.
- 5 recordings / month
- All 4 asset types
- Up to 5 minutes per recording
- Browser-only transcription
- "Made with Cassette" footer always on
Creator
Unlimited recordings. Save your library. Pick your voice.
- Unlimited recordings
- Whisper-quality transcription (coming soon)
- Tone presets — sharp, warm, founder-voice
- Save & search your asset library
- Toggle the footer off
- Export to Markdown / Notion / Beehiiv
Studio
Three seats. Shared library. Brand voice training.
- Everything in Creator
- 3 seats included
- Shared library + assignments
- Custom brand voice training
- API access (coming soon)
- Priority support
The honest answers.
Does this actually sound like me, or like generic AI?
Cassette is built around one rule: keep the speaker's voice. The generation prompt explicitly preserves your vocabulary, your rhythm, and your specific examples. If you say "absolute graveyard" we don't change it to "underutilized resource." If you mention a specific number, we keep it; we never invent one.
How long can a recording be?
Free plan: 5 minutes per recording. Creator and Studio: 30 minutes. The sweet spot for great output is 60 seconds to 4 minutes — long enough to develop a thought, short enough to stay sharp.
Where does my audio go?
On v1, transcription happens in your browser using the native SpeechRecognition API — your audio never leaves the device. Only the transcript is sent to our servers to generate assets, and we don't store transcripts on Free plans. Whisper-quality cloud transcription (an opt-in upgrade) is on the Day-30 roadmap.
What if the assets aren't quite right?
Today, copy what works and edit what doesn't — Cassette is a 70%-of-the-way-there tool, not an autopilot. Tone presets, regenerate, and per-asset feedback ("more punchy", "less corporate") are on the immediate roadmap.
Why a tape cassette?
Because every great voice memo deserves to be treated like Side A — not an "untitled_001.m4a" in a buried folder.
Who's behind this?
Cassette is a product of RheaForge, the build engine inside Twin Flame Group. We ship one production-grade product per day. Cassette was Day 001b.
The voice memo you'd publish if it took ten seconds.
Cassette makes it take ten seconds.
Hit record →No signup. Works on iPhone, Android, and desktop browsers.