Precision Transcription
Speaker-aware transcription with timestamps. Upload audio or record live — every word captured with professional-grade accuracy. 99+ languages supported.
Private by design — EU/GDPR · Swiss ISO 27001
Every word.
On the record.
AI transcription and meeting intelligence for professionals who cannot afford to get it wrong.
99.1%
accuracy
< 2 min
per hour of audio
0
audio stored by us
14 days
free trial
Built for professionals on the record
BigLaw & LitigationHealthcare & MedicineResearch & AcademiaGovernment & PolicyExecutive Teams
Product
Everything you need.
Nothing you don’t.
Meeting Intelligence
Automatic summaries, action items, and key decisions extracted by AI. Turn an hour-long recording into a structured brief in under 2 minutes.
Chat with Your Meetings
Ask questions across one or many recordings. "What did the client say about the deadline?" Answered instantly — only from what was actually said.
Privacy by Architecture
Audio never leaves your machine. Everything runs locally — no cloud, no third parties, no data sharing. No training on your data. Ever.
Searchable Archive
Every meeting stored, indexed, and searchable. Find the exact moment a decision was made across months of recordings.
Jurisdiction Choice
US-based? EU compliance rules? Swiss regulations? You choose where text is processed. Your audio never leaves your device on the Private tier.
How It Works
Set up once.
Private forever.
OneTranscript downloads AI models to your device during first-time setup. After that, everything runs locally — no internet, no cloud, no data leaving your machine.
01
Download the app
Download OneTranscript for macOS, Windows, or Linux. The app is about 100 MB — a quick download on any connection.
02
One-time AI setup
On first launch, OneTranscript downloads open-source AI models (~5 GB) to your device. This takes 5-15 minutes and is the only time an internet connection is needed. You can use your computer normally while it downloads.
03
Record or upload — fully offline
Record meetings live or upload audio files. Speech recognition and AI summarization run entirely on your device. No audio is sent to any server — not to OneTranscript, not to the AI model providers, not to anyone.
04
Chat, search, and act
Ask questions about your meetings, search across your archive, extract action items, and generate follow-ups. All powered by local AI. Your data stays encrypted on your device — private forever.
Pricing
Simple, honest pricing.
No feature gating.
14-day free trial. Card required — cancel anytime. A lawyer billing $400/hr saves 2+ hours/month. OneTranscript pays for itself in 2 minutes.
Desktop App
Private
Audio never leaves your machine
$12/mo
$96/yr — save $48
- On-device transcription — audio never leaves your machine
- AI summaries, action items, and meeting intelligence
- Chat with your meetings (temperature-anchored, no hallucination)
- Searchable transcript archive
- 14 language support with auto-detection
- macOS, Windows, and Linux desktop app
- Offline-capable — works without internet after setup
- AES-256-GCM encrypted local database
System requirements
RAM8 GB min · 16 GB recommended
Storage10 GB free (models download on first run)
OSmacOS 12+ · Windows 10/11 · Ubuntu 20.04+
CPUIntel i5 2018+ · AMD Ryzen 5 · Apple M1–M5
Speed60 min recording: Apple M2 ~5 min · Intel i7 ~20 min
Self-Hosted
Enterprise
Your infrastructure. Your control.
Custom
From $5k/yr · volume discounts
- Docker self-hosted — air-gapped, zero external dependencies
- SSO via SAML / OIDC (Okta, Azure AD, Google Workspace)
- Custom DPA + BAA for HIPAA-aligned deployments
- Audit logs and compliance reporting
- Self-hosted speaker diarization — no third-party audio processing
- Unlimited seats and usage
- SLA with guaranteed uptime
- Dedicated account manager and implementation support
No feature gating · No usage caps · Cancel anytime · 14-day full trial
Use Cases
Purpose-built for your profession.
Legal
On the record. Every time.
Depositions, client calls, partner meetings. Every word attributed to the right speaker, timestamped and searchable. Privilege preserved — EU and Swiss options for cross-border matters.
- Deposition notes
- Client call records
- Partner meeting intelligence
- Cross-border GDPR compliance
Medicine
Patient notes. Without the paperwork.
Record patient consultations and rounds. AI generates structured notes so physicians spend time with patients, not keyboards. HIPAA-aligned hosting options available.
- Patient consultation notes
- Rounds summaries
- Referral letters
- HIPAA-aligned options
Research
Every insight, captured.
Research interviews, focus groups, field recordings. Multi-speaker transcription with timestamps. Chat across your entire corpus to surface patterns no human could find alone.
- Research interview transcription
- Focus group analysis
- Lecture capture
- Cross-corpus AI search
FAQ
Common questions
Does OneTranscript send my audio to the cloud?
No. On the Private tier, all speech recognition and AI processing happens locally on your device. Your audio files, transcripts, and meeting data never leave your machine. No data is sent to OneTranscript servers, AI model providers, or any third party.
Do I need an internet connection to use OneTranscript?
Only for the initial setup. When you first install OneTranscript, it downloads open-source AI models (~5 GB) to your device. This takes 5-15 minutes depending on your internet speed. After that, OneTranscript works completely offline — no internet connection required.
Can OneTranscript or the AI model companies access my data?
No. OneTranscript Private runs entirely on your device. We have no servers that receive your data. The AI models (speech recognition and language models) run locally and make no network connections. Your transcripts are encrypted with AES-256-GCM on your device. No one — including OneTranscript — can access your recordings or meeting data.
What languages does OneTranscript support?
OneTranscript supports 14 languages with automatic language detection: English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Chinese, Japanese, Korean, Arabic, Hindi, and Turkish. Additional languages are added regularly.
What are the system requirements?
OneTranscript requires macOS 12+, Windows 10/11, or Ubuntu 20.04+. Minimum 8 GB RAM (16 GB recommended), 10 GB free disk space, and a modern CPU (Intel i5 2018+, AMD Ryzen 5, or Apple M1-M5). Apple Silicon Macs get 5-15x faster transcription speeds.
How is OneTranscript different from Otter.ai, Fireflies, or other cloud transcription services?
Unlike cloud-based services, OneTranscript Private processes everything locally on your device. Your audio never touches a server. This makes it suitable for confidential conversations — legal depositions, medical consultations, executive meetings — where cloud processing would create compliance and privacy risks. No subscription to a cloud service is required after the initial purchase.
Early Access
The future of professional transcription is private.
Join the waitlist. Be first when we open the doors.
No spam. Unsubscribe anytime.