Quickstart

If your code already talks to the Anthropic Messages API, change two things: the base URL and the key.

Endpoint

Base URL	`https://api.nightferry.io`
Path	`POST /v1/messages`
Auth header	`x-api-key: <your key>`
Version header	`anthropic-version: 2023-06-01`

A request

curl https://api.nightferry.io/v1/messages \
  -H "x-api-key: $NIGHTFERRY_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-opus-4-6",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Streaming

Set "stream": true. Responses are server-sent events. They are not token by token: a 140-token answer typically arrives as two or three chunks after about three seconds. Streaming gives you the same total latency as a normal request, so use it to start rendering early — not to reduce time to first token.

curl -N https://api.nightferry.io/v1/messages \
  -H "x-api-key: $NIGHTFERRY_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{"model":"claude-sonnet-4-6","max_tokens":1024,"stream":true,
       "messages":[{"role":"user","content":"Hello"}]}'

Set a generous client timeout. Long generations have been observed to take well over a minute. A timeout that fires early does not cancel the work — see Errors.

Desktop clients

For Cherry Studio, CC Switch, and similar tools, choose the Anthropic provider and override the base URL to https://api.nightferry.io, then paste your key in the API key field. No other change is needed.

Models

GET /v1/models returns everything your key can reach. Current rates are on the pricing page.