Overview

Console

Keys, usage and the endpoints to call. The dashboard comes first, the details live one tab over.

Go deeper

The other tabs this console keeps, one click from the same glance.

Recent requests

Written after every reply lands, with its model and token count.

Endpoint

Point any OpenAI client here.

Base URL

OpenAI clients take this verbatim as base_url. Endpoints: /chat/completions and /models.

https://chat.merkium.nl/api/v1

Rate limit

One request every two seconds per API key. Over the line gets a 429 with Retry-After. The window is per key, so separate keys keep moving.

First request

Streaming, reasoning and tool calls come back exactly as OpenAI sends them.

curl https://chat.merkium.nl/api/v1/chat/completions \
  -H "Authorization: Bearer $MERKIUM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"ganymede-1.1","messages":[{"role":"user","content":"Hi"}]}'