Skip to Content
API ReferenceLLM Provider Schema

LLM Provider Configuration Schema

Model: LlmSetting

Fields:

FieldTypeRequiredDescription
providerenumYesopenai, anthropic, google, openai-compatible
modelstringYesExact model identifier from the provider
api_keystringYesAPI key from the provider’s developer console
base_urlstringNoRequired only for openai-compatible. API endpoint URL.
is_activebooleanYesOnly one setting can be active at a time

Provider-Specific Notes:

  • OpenAI: Use models like gpt-4o, gpt-4-turbo. No base_url needed.
  • Anthropic: Use models like claude-3-5-sonnet-20241022. No base_url needed.
  • Google: Use models like gemini-1.5-pro. No base_url needed.
  • OpenAI-compatible: Requires base_url (e.g., <your-provider-endpoint>). Supports DeepSeek, Qwen, Together, OpenRouter, Groq, Ollama, and any OpenAI-compatible API.

Test Connection Endpoint:

POST /super-admin/llm/test

Rate limited to 5 requests per minute. Sends a minimal structured prompt to verify the provider responds correctly.

Last updated on
Webhook EndpointsEnvironment Variables