FAQ

FAQ

How do I install?

Python 3.11+, optional Ollama. Clone the repo, create a venv, pip install -e ., then python -m voxoryl. Or grab a Win/Mac zip from Releases.

Step-by-step: README → Quick start.

How do I wake it?

Say “Hey Voxy” (also “Hey Voxoryl” / “Okay Voxy”). You’ll see the floating voice orb. Realtime stop/cancel/pause commands bypass the LLM.

Where does my data live?

Personal files stay out of the git clone by default:

OS	Root
Windows	%LOCALAPPDATA%\VOXORYL (usually C:\Users\<you>\AppData\Local\VOXORYL)
macOS	~/Library/Application Support/VOXORYL/
Linux	~/.local/share/voxoryl/

Inside that folder: config.env, data/ (memory, knowledge, reels, logs), and a workspace sandbox. Override with VOXORYL_USER_DATA_DIR if you need to.

Contributors can set VOXORYL_USE_REPO_DATA=1 to keep data next to the repo (legacy / dev).

Windows vs Mac?

Both run the API, voice widget, memory, and Doctor. Native orb is strongest on Windows (WebView2). Computer-use (screen click/type) is Windows-first when enabled; Mac is best-effort / limited. Chrome & YouTube work on both, with Windows the stronger path for coordinate control.

Is it cloud-only?

No. Default reasoning is on-device via Ollama. Optional cloud (Groq / Gemini / etc.) is for general chat — tools still run on your PC. Sensitive flows stay local-only.

Something broke — what now?

Use Doctor / Safe Mode from the product UI or docs in the README. Prefer filing a real bug report with OS + steps.