Start in five minutes

1. Get your key

Create an account, verify your email and your card, and copy the key from the panel. It is shown once.

2. Change the base URL

Wherever your tool asks for OpenAI's URL, put ours. Nothing else about the request changes.

https://api.enroutia.com/v1

3. Make the call

That is all of it. If your automation worked with OpenAI, it works the same way now.

curl https://api.enroutia.com/v1/chat/completions \
  -H "Authorization: Bearer TU_CLAVE" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "auto",
    "messages": [{"role": "user", "content": "Di hola en una frase."}]
  }'

Which model to ask for

Ask for a specific model from the catalogue, or ask for “auto” and we keep it pointed at the best balance of quality and price. Ask for a name that does not exist and you do not get an error: we answer with “auto” and tell you so in the X-Resolved-Model header.

Or a name of your own

Create an alias in the panel — `enroutia/keywords-prod` — and put that in the model field instead. Your automation keeps the name; you change the model behind it from the panel whenever you like, with a history and a rollback.

{"model": "enroutia/keywords-prod", "messages": [...]}

Streaming

Add stream: true and the answer arrives in pieces, in the same SSE format as OpenAI, final [DONE] included.

{"model": "auto", "stream": true, "messages": [...]}

Not on chat completions?

The gateway also answers OpenAI's Responses API and Anthropic's Messages API on the same base URL and the same key, so an SDK written against either works with the base URL change alone. Responses API and Anthropic SDK

POST https://api.enroutia.com/v1/responses   {"model": "auto", "input": "Di hola en una frase."}
POST https://api.enroutia.com/v1/messages    {"model": "auto", "max_tokens": 256, "messages": [...]}