Does it work without an internet connection?

Yes. Transcription runs entirely on your machine using the local model you've downloaded. Once the model is cached, no network connection is required.

What do I get when I buy?

You get compiled installers for all 3 platforms: Linux (.AppImage), macOS (.dmg), and Windows (.exe). No Python installation required — just download and run.

What's the difference between Free and Pro?

Free includes the base model and 7-day trial with all features. Pro ($19/mo or $190/yr) unlocks all local models, unlimited history, dictionary persistence, full stats, and priority support.

Settings and Customization

Access the Settings window by right-clicking the Agent Wispr widget and selecting Settings.... A first-run walkthrough now guides you through microphone setup before model selection.

General Tab

Hotkey: Click the record button and press your desired key combination (e.g., Alt+Space).
Mode: Switch between Push-to-Talk and Toggle recording behaviors.
Theme: Choose between Dark, Light, or System appearance.
Accent Color: Change the primary color of the widget and waveform.
Opacity: Adjust how transparent the widget appears when not in use.
Always on Top: Keep the widget visible above all other windows.

Model Tab

Wispr Model: Choose the model size that balances speed and accuracy for your hardware.
- tiny: Fastest, lowest resource usage.
- base: Good balance for most dictation.
- small: Improved accuracy.
- medium / large-v3: Best accuracy (Pro only).
Processing Device: Select Auto (recommended), CPU, or CUDA (requires NVIDIA GPU).
Language: Force transcription to a specific language or leave as Auto for detection.

Microphone Tab

Microphone Device: Select the input device Agent Wispr should record from.
Refresh: Reload the available microphones if you connect or change hardware after launch.
Test: Run a local microphone test to confirm capture and view the signal level.
macOS Permissions: If the mic is silent on Mac, check System Settings > Privacy & Security > Microphone and make sure Agent Wispr is allowed.
Linux / WSL: If no devices appear, confirm your operating system is exposing an input device before launching the app.

AI Cleanup Tab (Pro)

Enable secondary LLM processing to polish your transcriptions.

Groq API Key: Required for cleanup. Get a free key at groq.com.
Cleanup Prompt: Customize the system instructions used to process your text.

Dictionary Tab

Manage your custom word corrections.

Add Row: Map a misheard word to its correct spelling.
Import/Export: Move your dictionary between devices via CSV files.

History Tab

Transcription Log: View, search, and re-type previous transcriptions.
Export: Save your history to a CSV file for analysis.
Clear All: Permanently delete your local transcription history.

About Tab

Edition: Shows your current edition -- Free Edition or Pro Edition.
License Key: Paste your license key here and click Activate to unlock Pro features. See Activating Pro for step-by-step instructions.

General Tab

Hotkey: Click the record button and press your desired key combination (e.g., Alt+Space).

Mode: Switch between Push-to-Talk and Toggle recording behaviors.

Theme: Choose between Dark, Light, or System appearance.

Accent Color: Change the primary color of the widget and waveform.

Opacity: Adjust how transparent the widget appears when not in use.

Always on Top: Keep the widget visible above all other windows.

Model Tab

Wispr Model: Choose the model size that balances speed and accuracy for your hardware.

tiny: Fastest, lowest resource usage.
base: Good balance for most dictation.
small: Improved accuracy.
medium / large-v3: Best accuracy (Pro only).

Processing Device: Select Auto (recommended), CPU, or CUDA (requires NVIDIA GPU).

Language: Force transcription to a specific language or leave as Auto for detection.

Microphone Tab

Microphone Device: Select the input device Agent Wispr should record from.

Refresh: Reload the available microphones if you connect or change hardware after launch.

Test: Run a local microphone test to confirm capture and view the signal level.

macOS Permissions: If the mic is silent on Mac, check System Settings > Privacy & Security > Microphone and make sure Agent Wispr is allowed.

Linux / WSL: If no devices appear, confirm your operating system is exposing an input device before launching the app.