FAQ

Frequently asked questions about Trueears

General

What is Trueears?

Trueears is a minimalist, context-aware AI voice dictation application. It transforms speech into intelligently formatted text based on the active application — for example, producing casual Slack messages, structured code comments in VS Code, or formal email copy in Outlook.

How much does it cost?

Trueears itself is free and open source. You only pay for API usage:

• Groq API (transcription): free tier available with generous limits
• LLM post-processing (formatting): uses Groq’s free chat models

No subscription or license fee is required to run Trueears.

Which platforms are supported?

Platform	Status
Windows 10/11	Fully supported
macOS	Coming soon
Linux	Coming soon

---

Privacy & security

Does Trueears store my audio?

No. Audio is:

1. Recorded locally in memory
2. Sent to Groq for transcription
3. Immediately discarded after transcription completes

Nothing is persisted to disk or to any server you do not control.

Where are my API keys stored?

API keys (Groq, OpenRouter) are stored in Tauri’s Store plugin — a settings.json file in the app data directory. Authentication tokens from Google OAuth are stored in a separate auth.json file in the same directory.

Keys are never stored in localStorage or sent to any server other than the APIs they authenticate with.

Is my voice data sent anywhere besides Groq?

No. Audio is only sent to Groq’s Whisper API for transcription. If you enable LLM post-processing, the transcribed text (not the audio) is sent to Groq’s chat API for formatting. No other third-party services receive your data.

Can I use Trueears offline?

Not currently. Transcription requires a live internet connection to the Groq API. Local and offline model support is planned for a future release.

---

Features

What's the difference between recording modes?

Mode	Behavior	Best for
Auto (default)	Quick tap = Toggle, Hold = Push-to-Talk	Flexibility
Toggle	Press to start, press again to stop	Long dictation
Push-to-Talk	Hold to record, release to stop	Quick commands

Configure the mode in Settings > Preferences.

What is Select-to-Transform?

Select-to-Transform lets you rewrite existing text using a voice instruction:

1. Select text

   Highlight any text in any application.
2. Activate recording

   Press Ctrl+Shift+K.
3. Speak your instruction

   Say a transformation, for example: “make it professional”, “translate to Spanish”, or “convert to bullet points”.
4. Stop recording

   Press Ctrl+Shift+K again. The selected text is replaced with the transformed version.

What is Log Mode?

Log Mode saves voice notes as timestamped entries in app-specific markdown files.

Speak a trigger phrase followed by your note:

Trigger phrase	Example
Log	”Log fix the login button”
Remember	”Remember to update the docs”
Note to self	”Note to self check the API limits”

Entries are appended in the following format:

- [2025-12-29 14:30] Fix the authentication bug
- [2025-12-29 15:45] Review PR #42 before merging

Configure Log Mode in Settings > Log Mode.

Can I customize app profiles?

Yes. Go to Settings > App Profiles to:

• Add profiles for new applications
• Modify formatting instructions per app
• Set context-specific system prompts
• Override the transcription language per app

Can I use different languages for different apps?

Yes. Each App Profile supports a Language Override:

1. Open App Profiles

   Go to Settings > App Profiles.
2. Create or edit a profile

   Select an existing profile or click Add Profile.
3. Set the language

   Under Language Override, click Change, select your preferred language, and click Save.

The per-profile language takes priority over the global transcription language setting. This is useful for multilingual workflows — for example, Spanish for WhatsApp and English for VS Code.

---

Technical

Why Tauri instead of Electron?

	Tauri	Electron
Bundle size	~15 MB	~150 MB
Memory usage	Lower	Higher
Backend	Native Rust	Node.js
WebView	System WebView	Bundled Chromium

Tauri produces a roughly 10x smaller bundle and lower memory overhead by using the operating system’s built-in WebView rather than bundling a full browser engine.

What LLM models are supported?

Trueears currently supports OpenRouter-hosted models for LLM post-processing:

• openai/gpt-oss-120b (recommended)
• openai/gpt-oss-20b

Custom endpoint support is planned for a future release.

What Whisper model is used for transcription?

Trueears uses whisper-large-v3-turbo via the Groq API by default. This model provides a good balance of transcription speed and accuracy.

You can change the model in Settings > Transcription.

How do global hotkeys work?

Trueears registers system-wide hotkeys using Tauri’s global-shortcut plugin:

Action	Shortcut
Start / stop recording	Ctrl+Shift+K
Open settings	Ctrl+Shift+S

These shortcuts are active even when Trueears is not the focused window. Note that elevated processes (applications running as administrator) may block global hotkeys from being received by Trueears.