Before you connect#
AI access requires sign-in, one confirmed top-up of at least USD 100, an authorized API key and enough available credit for the request. Each later top-up also has a USD 100 minimum; smaller remaining balances stay usable when they cover the request. Credit requires transaction verification. Check service status for model availability before sending traffic. Use your own API key without the Bearer prefix in key fields. The base URL already includes /v1. These configurations were checked against official documentation, not tested against a live GPU deployment.
Configuration#
Obsidian itself has no universal LLM API settings. Name the plugin explicitly: Copilot for Obsidian. Open Settings > Copilot > Models (BYOK) and add a 3rd Party (OpenAI-Format) provider/custom model. Set Base URL to https://weightsapi.com/v1, the API key to the service key, and model name to Qwen/Qwen3-32B. Save and enable it under Basic > Agents > Quick Chat models. Older Copilot versions call the action Add Custom Model; the custom form has model name, provider, Base URL and API key.
If direct requests fail with CORS, Copilot documents a CORS workaround toggle, with streaming unavailable in that mode. This chat-model recipe does not configure vault embeddings; use a separately supported embedding model for vault indexing/RAG.
Sources: Current Copilot provider docs, custom model form examples.
Verify the connection#
Fetch the model list, then send one short message. A successful model list does not prove tools, vision or JSON schema support. Check the selected deployment’s capabilities before enabling them.