Here’s how you can make your first API request using the ShuttleAI API. Replace $SHUTTLEAI_API_KEY with your API key, which you can find on our Dashboard.

Each model has a specific cost that equates to the number of “requests” deducted per use. Find the cost and limits for each model on your Dashboard and detailed model cost information here.
curl https://api.shuttleai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $SHUTTLEAI_API_KEY" \
  -d '{
     "model": "shuttle-3",
     "messages": [{"role": "user", "content": "Say this is a test!"}],
     "temperature": 0.7,
     "max_tokens": 5
   }'

When you send a request to the shuttle-3 model with specified parameters like max_tokens and temperature, you’ll receive a response akin to this:

{
  "id": "chatcmpl-6ee59e89dfd1436cbf29d90c3c2168f5",
  "object": "chat.completion",
  "created": 1730229481,
  "model": "shuttle-3",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! How can I assist you today?"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 7,
    "completion_tokens": 9,
    "total_charged": 9.8e-6
  }
}

This response illustrates the successful processing of your ChatCompletion request.

​
Understanding the Response

  • finish_reason: “length” indicates that the response ended because it reached the max_tokens limit. Other possible values include “stop” and “tool_calls”.
  • completion_tokens: Shows the number of tokens used to generate the response, which will not exceed the set max_tokens.

Want to receive data in real-time as it’s generated?

Explore Streaming!