# Thoth

> Private AI meeting scribe for Mac, iPhone, and iPad. Records both sides of any meeting, then transcribes, identifies speakers, and summarizes locally on your device. No cloud, no data leaving your machine.

Thoth is a native Apple app that captures your microphone and system audio on separate channels, transcribes everything using on-device AI (WhisperKit, CoreML, Parakeet, Apple Speech), identifies who spoke when, and generates AI summaries without uploading your audio anywhere. Cloud AI is optional and opt-in, and on-device anonymization can redact names, places, phone numbers, and emails before any text leaves the device. Built by one developer. Available on the Mac App Store, with iPhone and iPad in open beta on TestFlight.

The macOS app requires macOS Tahoe (macOS 26) or later. Apple Silicon is recommended for on-device AI summaries. Intel Macs are supported for recording and transcription. The app is fully localized in English and French.

## AI Usage

The marketing content on this site may be read, indexed, summarized, and cited by AI assistants and crawlers. We welcome it: robots.txt explicitly allows GPTBot, ClaudeBot, Google-Extended, and PerplexityBot. The Thoth app itself processes user audio, transcripts, and summaries entirely on device by default. Cloud AI is optional, opt-in, and supports bring-your-own-key with privacy-focused providers; on-device redaction can mask personal data before any transcript is sent. No user recordings, transcripts, or summaries are ever used to train any model.

## Features

- On-device transcription with multiple engines (WhisperKit, Parakeet, Apple Speech) and selectable model sizes
- Speaker identification (diarization) that color-codes who said what (Pro)
- On-device anonymization that redacts names, places, phone numbers, and emails before optional cloud AI
- Local AI summaries, plus optional bring-your-own-key cloud summaries with privacy-focused providers
- Saved custom prompts to rerun your own AI instructions on any recording
- Separate microphone and system-audio capture, mixed or independent
- Live Activity and Dynamic Island for active recordings on iPhone
- Direct device-to-device transfer of recordings between iPhone and Mac over local WiFi or AirDrop
- Export to M4A, AAC, Markdown, RTF, JSON, and PDF

## Key Pages

- [Homepage](https://thoth-app.com/): Product overview, features, benchmarks, pricing
- [French Homepage](https://thoth-app.com/fr/): The full site in French
- [Blog](https://thoth-app.com/blog/): Technical posts on transcription, AI, and indie Mac development
- [Transcription Accuracy](https://thoth-app.com/transcription-accuracy/): Full benchmark results across all models
- [Alternatives](https://thoth-app.com/alternatives/): How Thoth compares to Otter, Fireflies, Granola, MacWhisper
- [Granola Alternative](https://thoth-app.com/granola-alternative/): Thoth vs Granola
- [Otter Alternative](https://thoth-app.com/otter-alternative/): Thoth vs Otter.ai
- [Fireflies Alternative](https://thoth-app.com/fireflies-alternative/): Thoth vs Fireflies.ai
- [MacWhisper Alternative](https://thoth-app.com/macwhisper-alternative/): Thoth vs MacWhisper
- [Use Cases](https://thoth-app.com/use-cases/): Who uses Thoth and why
- [Use Case: Lawyers](https://thoth-app.com/use-cases/lawyers/): Client meetings, depositions, privileged conversations
- [Use Case: Journalists](https://thoth-app.com/use-cases/journalists/): Source interviews, off-the-record protection
- [Use Case: Doctors](https://thoth-app.com/use-cases/doctors/): Patient consultations, HIPAA considerations
- [Use Case: Researchers](https://thoth-app.com/use-cases/researchers/): Field interviews, qualitative data
- [Use Case: Students](https://thoth-app.com/use-cases/students/): Lectures, study notes, recall
- [Other Apps](https://thoth-app.com/apps/): Other apps by the developer
- [Roadmap](https://thoth-app.com/roadmap/): Planned features
- [Press](https://thoth-app.com/press): Press kit and media resources
- [Support](https://thoth-app.com/support/): Help and contact

## Blog Posts

- [Thoth is coming to iPhone](https://thoth-app.com/blog/2026-06-05-thoth-for-iphone/): The private AI scribe for Mac is coming to iPhone and iPad. Record, transcribe, diarize, and summarize on device, beta open now on TestFlight.
- [Redact names before cloud AI sees them](https://thoth-app.com/blog/2026-06-05-on-device-anonymization/): Thoth finds and masks names, places, phone numbers, and emails on device, then swaps them for tokens before any cloud AI sees your transcript.
- [Thoth for ADHD and autism: what I learned](https://thoth-app.com/blog/2026-05-22-thoth-adhd-autism-accessibility/): I did not build Thoth thinking about accessibility. A review made me realize I should have. This is a subject that matters to me personally.
- [A band-aid for Parakeet's language drift](https://thoth-app.com/blog/2026-05-20-fixing-parakeet-language-drift/): Parakeet TDT v3 drifts into English on spontaneous French speech. A workaround using the top-64 logits: not a fix, but good enough to change the recommendation.
- [Parakeet translates French audio](https://thoth-app.com/blog/2026-05-19-parakeet-language-drift/): We added Parakeet TDT v3 to Thoth expecting fast transcription. On non-English audio, it was doing something stranger than we anticipated.
- [Thoth is coming to your menu bar](https://thoth-app.com/blog/2026-05-18-menu-bar-mode/): A first look at Thoth's menu bar mode: one-click recording, live transcript, and a floating window that stays on top across every Space and fullscreen app.
- [Could Thoth come to iPhone? What I found](https://thoth-app.com/blog/2026-05-18-thoth-ios-exploration/): I spent time mapping out what a Thoth iOS app would actually look like. Here is what works, what does not, and the ideas that surprised me.
- [Your meeting audio belongs on your Mac](https://thoth-app.com/blog/2026-05-13-why-your-meeting-recorder-shouldnt-upload-your-audio/): Cloud meeting recorders upload your audio to US servers, creating legal exposure for everyone on your calls. Here is what that means in practice.
- [Local vs cloud AI summaries: a benchmark](https://thoth-app.com/blog/2026-05-13-local-vs-cloud-ai-summaries/): I tested Phi 3.5 Mini, Gemma 3 12B, and Claude Sonnet across two real meetings with three independent judges. Here is what I found.
- [How I benchmark transcription models in Thoth](https://thoth-app.com/blog/2026-05-10-how-we-benchmark-transcription/): I ran three real-world scripts through every model in Thoth and measured Word Error Rate. Here's what I found.

## Pricing

- Free: Unlimited recordings, 30 min per mic recording, 15 min system audio, 10 AI enhancements per month
- Pro Monthly: $9.99/month
- Pro Yearly: $49.99/year
- Lifetime: $99.99 one-time (EUR 99.99)

Pro unlocks: unlimited recording duration, system audio and mixed recording, speaker identification, all export formats (M4A, AAC, Markdown, RTF, JSON, PDF), unlimited AI enhancements, large transcription model, Parakeet models. One Pro purchase unlocks every platform.

## Download

- Mac App Store: https://apps.apple.com/app/thoth-your-private-ai-scribe/id6756965313?mt=12
- iPhone and iPad (open beta): https://testflight.apple.com/join/hu2xjnN7