Making Requests to the i14 API: Chat Completions Guide

The i14 API is OpenAI-compatible. If you’ve used the OpenAI API, the request format will feel familiar - only the base URL and model ID are different. You can migrate existing OpenAI integrations by swapping two values and be up and running immediately.

Base URL

All requests are made to the following base URL:

https://api.i14.ai/v1

To get your API key, see Get your API key.

Request Format

Send a POST request to /v1/chat/completions with your API key in the Authorization header and a JSON body containing the model, messages, and any optional parameters.

macOS and Linux (bash or zsh)

curl https://api.i14.ai/v1/chat/completions \
  -H "Authorization: Bearer $I14_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "i14-sugarglider-nemotron3-super-120B-v1",
    "messages": [
      { "role": "system", "content": "You are a helpful assistant." },
      { "role": "user", "content": "What is the capital of France?" }
    ],
    "temperature": 0.7,
    "max_tokens": 2048
  }'

Key Parameters

modelstringRequired

The model ID to use for the completion. Use i14-sugarglider-nemotron3-super-120B-v1.

messagesarrayRequired

An array of message objects that form the conversation history. Each object must include a role (system, user, or assistant) and a content string. Sugar Glider (i14-sugarglider-nemotron3-super-120B-v1) accepts text only.

temperaturenumber

Sampling temperature between 0 and 2. Higher values produce more creative output; lower values produce more deterministic output. Defaults to 1.

max_tokensinteger

The maximum number of tokens to generate in the response. If omitted, the API stops after 100 tokens and returns "finish_reason": "length", so set it explicitly. Sugar Glider is a thinking model, and thinking tokens count toward this limit. Use a value of 2048 or more so the answer is not cut off by the thinking.

streamboolean

Set to true to receive the response as a stream of Server-Sent Events (SSE). See the Streaming page for details.

top_pnumber

Nucleus sampling threshold. The model considers only the tokens comprising the top top_p probability mass. Defaults to 1.

Response

A successful non-streaming request returns a JSON object like the following:

{
  "id": "req_fbc98589b09740fd8eda2e28d6306ec8",
  "object": "chat.completion",
  "created": 1791426828,
  "model": "i14-sugarglider-nemotron3-super-120B-v1",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "\n\nThe capital of France is Paris. It is the country's political, cultural, and economic center, housing major government institutions such as the Élysée Palace (the presidential residence) and the National Assembly. Paris has been the capital since the Middle Ages and remains a global hub for art, fashion, and history.",
        "reasoning_content": "Okay, the user is asking for the capital of France. Let me start by re..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 44,
    "completion_tokens": 369,
    "total_tokens": 413,
    "completion_tokens_details": {
      "reasoning_tokens": 305
    }
  },
  "x_i14_perf": {
    "tokens_per_second": 36.227,
    "generation_time_s": 10.1859,
    "time_to_first_token_s": 0.1052
  }
}

The choices[0].message.content field contains the model’s answer. The choices[0].message.reasoning_content field contains the thinking that came before it. The example is trimmed: the real reasoning_content is longer. The usage object reports token counts for the request, which you can use for cost tracking. The i14-specific x_i14_perf object reports generation speed and timing; it contains additional fields beyond the three shown here.

Using the OpenAI SDK

Because the i14 API is OpenAI-compatible, you can use the official OpenAI SDK in any language by pointing it at the i14 base URL. Pass your i14 API key as the api_key argument.

Python

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["I14_API_KEY"],
    base_url="https://api.i14.ai/v1",
)

response = client.chat.completions.create(
    model="i14-sugarglider-nemotron3-super-120B-v1",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is the capital of France?"},
    ],
    temperature=0.7,
    max_tokens=2048,
)

print(response.choices[0].message.content)

Node.js

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.I14_API_KEY,
  baseURL: "https://api.i14.ai/v1",
});

const response = await client.chat.completions.create({
  model: "i14-sugarglider-nemotron3-super-120B-v1",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "What is the capital of France?" },
  ],
  temperature: 0.7,
  max_tokens: 2048,
});

console.log(response.choices[0].message.content);