SillyTavern with LivRouter
SillyTavern can connect to LivRouter when you use its OpenAI-compatible connection mode. This lets you use supported Claude, GPT, DeepSeek, GLM, and other catalog models from your own Tavern setup with one LivRouter API key.
Connection settings
- Open SillyTavern's API connection or custom OpenAI-compatible provider settings.
- Choose the OpenAI-compatible connection type.
- Set the API URL to
https://api.livrouter.com/v1. - Enter your LivRouter API key.
- Choose a model ID from the current model catalog.
Use the OpenAI-compatible endpoint:
https://api.livrouter.com/v1. Do not use the Anthropic base URL for SillyTavern's OpenAI-compatible connection mode.Example request
curl https://api.livrouter.com/v1/chat/completions \ -H "Authorization: Bearer $OPENAI_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "claude-sonnet-4-6", "messages": [ {"role": "user", "content": "Hello from my Tavern"} ] }'Choosing a model
Model IDs are deployment-specific and can change as the catalog is updated. If a model does not appear in SillyTavern, copy the exact ID from the LivRouter model catalogand refresh the provider model list.
Common issues
- 401 or unauthorized: check that the LivRouter API key is active and has available credits.
- Model not found: use the exact current model ID from the catalog.
- Requests fail immediately: confirm that the URL ends with
/v1and that the connection type is OpenAI-compatible.
For the full request format, see the API quickstart.