CostIQDocs

Start

Make your first request.

Connect to CostIQ from a terminal in a few minutes. You need a CostIQ-issued API key and cURL.

Before you begin

Ask your CostIQ administrator for a key created specifically for you or the service you are connecting. Do not share a credential across people or environments.

01

Configure your shell

The silent prompt keeps the key out of your terminal history and screen.

Shell
export COSTIQ_BASE_URL=https://inference.costiq.xyz/v1
export COSTIQ_MODEL=general-fast
printf 'CostIQ API key: ' >&2
read -r -s COSTIQ_API_KEY && printf '\n' >&2
export COSTIQ_API_KEY
02

List available models

A healthy, authorized request returns general-fast in the response data.

GET /models
curl "$COSTIQ_BASE_URL/models" \
  -H "Authorization: Bearer $COSTIQ_API_KEY"
03

Create a completion

Send messages using the familiar OpenAI Chat Completions shape.

POST /chat/completions
curl "$COSTIQ_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $COSTIQ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "general-fast",
    "messages": [
      {"role": "user", "content": "Explain this in one sentence."}
    ],
    "stream": false
  }'
04

Stream a response

Set stream to true and consume server-sent events until the terminal [DONE] event.

Streaming cURL
curl -N "$COSTIQ_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $COSTIQ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "general-fast",
    "messages": [{"role": "user", "content": "Write a haiku."}],
    "stream": true
  }'