Chatbox model setup: API keys and custom endpoints

This brief Chatbox walkthrough introduces the desktop and web clients, then concentrates on configuring a model provider. The presenter enters a request for an image-to-PDF web app in a new chat, but the recording does not establish a working generated application or compare the result with Claude.

The settings section covers selecting OpenAI or Gemini and entering the corresponding API key. It also shows a model URL field for a backend hosted by the user, and points to OpenRouter as another provider. These are connection settings: the video does not deploy a model server, download local weights or demonstrate offline inference. Hosted providers remain separate from a custom endpoint.

The presenter proceeds through email authorization before returning to the chat, then points to desktop and mobile downloads. The source provides a short introduction to account access and provider configuration; it does not substantiate the title's unlimited-use or competitive-performance claims.