Frequently Asked Questions
Everything you need to know about setting up, running, and customizing JARVIS for your workflow.
Is JARVIS completely free to use?
Yes, the core JARVIS software is 100% free and open-source under the MIT License. However, if you use cloud APIs like Groq, Gemini, or Regolo, you may incur standard API usage costs depending on their respective free tiers.
Can I run it completely offline?
Absolutely. You can point JARVIS to any local OpenAI-compatible endpoint. By running Ollama or LM Studio locally and updating the CUSTOM_BASE_URL in your .env, you can achieve 100% local, offline inference.
Does it work on Mac and Linux?
Currently, JARVIS is heavily optimized for Windows (native window controls, registry setups). Linux (.AppImage) and macOS (.dmg) versions are in active development and will be released very soon.
How does Mobile ADB control work?
It uses ADB (Android Debug Bridge) routed over Tailscale. You need to enable USB Debugging on your phone, run adb tcpip 5555, connect via your Tailscale IP, and add that IP to your .env.
Will it accidentally delete my files?
JARVIS employs a strict `shlex` tokenization and path canonicalization layer that auto-blocks highly destructive commands (like rm -rf /). Additionally, our Human-in-the-Loop (HITL) system ensures it asks for consent before major actions.
Why are WhatsApp and Telegram separated?
WhatsApp uses the Baileys Node.js web socket bridge for stable connections, requiring a QR code scan. Telegram uses the native MTProto API via Telethon for deeper OS integration and remote bot control. Both are managed safely as background services.
Still have questions? Check out the full Documentation or raise an issue on GitHub.