Messages (Anthropic) and Responses

Besides /v1/chat/completions, anthropic-format and Responses endpoints are available — for Claude Code and Codex CLI. All three share one base address and one key sk-pv-….

Authorize with the x-api-key: sk-pv-… header (as in the Anthropic API). The body is the Messages format: model, max_tokens, messages. SSE streaming is supported.

curl https://getvibeapi.com/v1/messages \
  -H "x-api-key: $GETVIBE_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-opus-5.5",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

The Responses endpoint (OpenAI Responses API format) — the path for Codex CLI. Authorization Authorization: Bearer sk-pv-….

Server-side conversation state is not supported: previous_response_iddoes not continue a prior request, and the store field is ignored. Send the full conversation history in the input field of every request.

curl https://getvibeapi.com/v1/responses \
  -H "Authorization: Bearer $GETVIBE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-6.1-sol",
    "input": "Hello!"
  }'

A stub endpoint for compatibility with the Anthropic SDK and Claude Code: they call it automatically before sending a request to estimate the input token count. Authorize with the same key sk-pv-… in x-api-key or Authorization: Bearer.

curl https://getvibeapi.com/v1/messages/count_tokens \
  -H "x-api-key: $GETVIBE_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-opus-5.5",
    "messages": [{"role": "user", "content": "Hello! Who are you?"}]
  }'

{ "input_tokens": 6 }