Skip to main content

Public API

Libre WebUI exposes an OpenAI-compatible API so existing SDKs, editors, and scripts can point at your instance with only a base-URL change.

Authentication​

Create a scoped API token under Settings → Account → API tokens (tokens use the lwk_ prefix and are stored hashed). The /v1 surface requires the chat scope. Pass the token as an ordinary bearer key:

export OPENAI_BASE_URL="https://your-instance/v1"
export OPENAI_API_KEY="lwk_..."

Requests are governed by the per-token rate limit (600 requests per minute) and the shared chat rate limit.

Endpoints​

EndpointPurpose
GET /v1/modelsAvailable models (local and provider-plugin backed)
POST /v1/chat/completionsStateless chat completion, streaming or not

POST /v1/chat/completions accepts the familiar body — model, messages (system/user/assistant; multimodal text parts are flattened), stream, temperature, top_p, max_tokens/max_completion_tokens, seed, and stop. Streaming responses are chat.completion.chunk SSE frames terminated by data: [DONE]; non-streaming responses include usage token counts when the provider reports them.

curl "$OPENAI_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama3.2",
    "messages": [{"role": "user", "content": "One sentence on herons."}]
  }'

Deliberate boundaries​

The /v1 surface is stateless inference: it never creates chats, touches your knowledge base, or runs tools. Stateful conversations with durable replayable events, cancellation by identity, idempotent enqueues, and the native tool loop live on the native /api/chat event API, which the same tokens can call with the chat scope. Roles other than system/user/assistant (for example tool) are rejected rather than silently accepted.