Home

AgentWhisper Wiki

AgentWhisper is push-to-talk voice dictation for Linux. Hold a key (F12 by default), speak, release — your words are transcribed on your own computer and typed into whatever you were writing, with a clipboard copy as backup. No cloud, no account, fully offline after the first model download.

Current version: v0.2.0 — deployable, in daily use.

Where to go

Page	What's there
Installation	Requirements, install, update, uninstall
Usage	Recording modes, the tray menu, the voice visualizer, CLI
Configuration	Every setting explained, model choices
Troubleshooting	Symptoms → causes → fixes
Architecture	How it works under the hood (for developers)
FAQ	Privacy, languages, Wayland, and other common questions

In one minute

sudo apt install python3-gi python3-gi-cairo gir1.2-ayatanaappindicator3-0.1 \
                 xclip xdotool libnotify-bin
git clone https://github.com/ChrisSchroedinger/agentwhisper.git
cd agentwhisper
curl -LsSf https://astral.sh/uv/install.sh | sh   # once, if uv is missing
./install.sh

Start AgentWhisper from your applications menu, wait for the model (first run only — check with agentwhisper status), then hold F12 and talk.