Chat

Chat with local or external OpenAI v1-compatible model providers. Provider URL and API key can be stored in local browser storage.

Provider

Choose a local preset or configure an external endpoint.

Connection

Requests are sent to the /chat/completions endpoint.

Provider URL

API Key(optional)

Remember API key in local storageOnly enable this on a trusted device. Local provider presets usually do not need a key.

Model

System Prompt

Chat

Send non-streaming OpenAI v1-compatible chat completions.

Start a conversation with llama.cpp.

Message

Press Ctrl+Enter or Cmd+Enter to send.