API Reference

Set the LLM provider

Configure and validate the active chat model.

POST/config/llmadmin

is_local: true targets a local Ollama instance; is_local: false requires provider_data.key and validates connectivity. Validation probes the exact provider_data.model given — an invalid or retired remote model is rejected with 400 and nothing is saved.

Body

is_localbooleanrequired

True for Ollama (local); false for remote providers.

provider_data.provider_nameenumrequired

openai | anthropic | ollama.

provider_data.modelstringrequired

The model id, e.g. qwen2.5-coder:latest or gpt-4o-mini.

provider_data.keystring

API key (empty for local providers).

Request

cURL — local Ollama

bash

curl -X POST localhost:8000/config/llm \
  -H "Authorization: Bearer $TOKEN" -H 'Content-Type: application/json' \
  -d '{"is_local":true,"provider_data":{"provider_name":"ollama","model":"qwen2.5-coder:latest"}}'

cURL — remote OpenAI

bash

curl -X POST localhost:8000/config/llm \
  -H "Authorization: Bearer $TOKEN" -H 'Content-Type: application/json' \
  -d '{"is_local":false,"provider_data":{"provider_name":"openai","model":"gpt-4o-mini","key":"sk-..."}}'

Responses

200Active configuration

json

{
  "is_local": true,
  "provider_name": "ollama",
  "model": "qwen2.5-coder:latest",
  "has_key": false,
  "base_url": null
}

400Invalid config / validation failed

json

{ "detail": "provider validation failed" }

← PreviousDelete a conversation Next →Get the current LLM config