Transcripts and notes that never leave your computer.
Spokenroom records or imports audio, transcribes it with word-level timestamps, labels who said what, and writes the summary — all as processes on your own Windows PC. No bot joins your call, no file is uploaded, no minutes are metered.
Built by a small independent team. No servers, and no investors asking us to monetize your recordings. Early-bird reservations are fully refundable until the release.
Interviews are given in confidence. Upload is a breach of that.
A source talks to you, not to a transcription vendor. A research participant signed a consent form that named your institution, not a cloud provider. A client under NDA expects the recording to stay inside the engagement. Every cloud transcription tool breaks those promises quietly, in the upload step.
Then there is the meter. Minutes caps, per-file limits, bots that join your calls and announce themselves, credits that vanish when you pause the subscription. You are paying rent on your own recordings.
The Mac has had a good one-time-purchase answer for years. Windows has not. This is that answer for Windows: the same speech models, running on the computer you already own, with speaker labels, timestamps, subtitles, and a summary template that cites its sources.
Record it or drop it in. Everything after that is local.
- Record or import.
Microphone, system audio (for calls), or both as separate tracks. Or drag in mp3, m4a, wav, mp4, mov and most other containers — a whole folder at a time if you like.
- Transcribe.
English runs fast on any modern CPU with word-level timestamps. Other languages use a multilingual model, best with a Vulkan GPU. Speakers are labelled, and you can rename or merge them in one click.
- Summarise.
The meeting-minutes template gives you key points, action items and quotations. Every quotation is checked against the transcript and links to its timestamp; anything that fails the check is dropped rather than softened.
- Export.
SRT and VTT for Premiere, DaVinci and YouTube. Timestamped transcript, DOCX, PDF, TXT, JSON. Click any word to jump the audio to it while you edit.
Nothing joins your call and announces itself. Nothing is uploaded, so there is nothing to cap. Transcribe a ten-minute voice memo or a six-hour oral history; the only limit is your disk.
Against the tool you use now, and the one your Mac friends use.
| Spokenroom | Otter.ai | MacWhisper | |
|---|---|---|---|
| Where audio is processed | Your computer | Vendor cloud | Your Mac |
| Price | $39 once | $16.99 / month, minutes capped | €64 once (Mac only) |
| Windows | Yes | Web | No |
| Bot joins your call | No | Yes, by default | No |
| Speaker labels | Yes; two-track for calls | Yes | Yes |
| Subtitles (SRT/VTT) | Yes | Limited | Yes |
| Summary cites its sources | Every quote links to a timestamp | No | No |
Competitor figures from public pricing pages, checked 2026-10-08. MacWhisper is a good product; if you are on a Mac today, buy it. This is for the rest of us. Therapist-edition comparison →
"Private" should mean something you can check.
Not a policy page. A list of the connections the software makes.
License check, which also works offline. Update check. Model download, only when you ask for one. That is the complete list.
There is no analytics SDK in the application. Crash reports are opt-in, anonymous, and contain a stack trace and nothing else.
Audio, transcripts and notes live in a directory you choose. Point it at an encrypted volume if your policy calls for that.
One switch blocks every outbound connection, and an indicator shows when it is on. The product is designed to be fully usable in that state.
When processing finishes, the recording is overwritten and removed unless you have chosen to keep it.
We have no server that ever receives your audio, so there is no jurisdiction whose request could reach it. The only thing that leaves your machine is a licence check you can switch off.
It runs on the laptop you already have.
Within reason, and we would rather tell you the limits now than refund you later.
- Runs
- Recommended
- Operating system
- Windows 10 or 11, 64-bit
- Windows 10 or 11, 64-bit
- Processor
- 4 cores with AVX2 (roughly 2013 onward)
- 8 cores recommended
- Memory
- 8 GB
- 16 GB
- Disk
- 5 GB
- 8 GB
- Graphics card
- Not required
- Not required; a Vulkan GPU makes notes faster and enables non-English transcription
One price, paid once.
- Unlimited recordings, unlimited minutes, forever
- One year of updates included; after that your version keeps working
- Optional update renewal at $29/year, never required
- Two machines per licence — office desktop and laptop
- Seven-day trial at release: every feature, no card, no account
- Refundable in full until the release, and for 30 days after
- Upgrade to the therapist edition any time for the $20 difference
Checkout opens when the merchant-of-record account is approved; until then, reserving puts you on the list for the early-bird price. Payment and invoicing are handled by our merchant of record, who also handles VAT and sales tax. We never see your card.
Hear when it ships — and get the launch discount.
One email a week at most: progress, benchmarks on real hardware, and what the models can and cannot do yet. No tracking pixels.