Get Started

Install and sign in

Download PandaOS, then choose where its models come from.

You can change this later

The provider is a setting, not a commitment: Settings, then Models. A single chat can also move between connections whenever you like.

Install

1. Download it from pandaos.ai/dashboard. — Pick your platform.
2. Run the installer and launch the app.
3. Answer the short onboarding. — Your name, how technical you want answers to be, and your AI provider.

Where the models come from

The one choice worth understanding. Seven routes:

• PandaOS Cloud, the default. Models from several providers, metered as usage, no keys to manage.
• Your API keys. Fourteen providers, plus any OpenAI-compatible endpoint. You choose the models and the billing.
• Local. Ollama, LM Studio, KoboldCpp and llama.cpp are found on their default ports.
• Free Mode. Cycles through the free tiers you already have with providers like Google, Groq and Mistral, moving on when one runs out.
• Claude, OpenAI Codex and OpenCode Go. Use a subscription you already pay for.