Back to Blog
EnglishTutorial

How to Access GPT-5 and GPT-5.2 via API - Complete Developer Guide

Learn how to access OpenAI's latest GPT-5, GPT-5.2, and o3-pro models through a unified API. Step-by-step guide with Python, Node.js, and curl examples.

C
Crazyrouter Team
January 22, 2026 / 1900 views
Share:
How to Access GPT-5 and GPT-5.2 via API - Complete Developer Guide

OpenAI has released its most powerful models yet: GPT-5, GPT-5.2, and the reasoning-focused o3-pro. This guide shows you how to access these cutting-edge models through Crazyrouter's unified API.

Supported OpenAI Models#

Crazyrouter provides access to the complete OpenAI model lineup:

ModelInput ($/1M tokens)Output ($/1M tokens)Best For
gpt-5.2$1.75$14.00Latest flagship, complex tasks
gpt-5.2-pro$3.50$28.00Enhanced reasoning
gpt-5$1.25$10.00General tasks
gpt-5-pro$2.50$20.00Advanced analysis
gpt-5-mini$0.25$2.00Cost-effective
gpt-5-nano$0.05$0.40High-volume tasks
o3-pro$20.00$80.00Complex reasoning
o3-mini$1.10$4.40Efficient reasoning
o4-mini$1.10$4.40Latest reasoning model

Quick Start#

1. Get Your API Key#

  1. Visit Crazyrouter Console
  2. Navigate to "Token Management"
  3. Click "Create Token"
  4. Copy your API key (starts with sk-)

2. Make Your First Request#

python
from openai import OpenAI

client = OpenAI(
    api_key="sk-your-api-key",
    base_url="https://crazyrouter.com/v1",
    default_headers={
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"
    }
)

response = client.chat.completions.create(
    model="gpt-5.2",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Explain quantum computing in simple terms."}
    ],
    temperature=0.7,
    max_tokens=1000
)

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

Using Node.js#

javascript
import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: 'sk-your-api-key',
  baseURL: 'https://crazyrouter.com/v1',
  defaultHeaders: {
    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
  }
});

async function main() {
  const response = await client.chat.completions.create({
    model: 'gpt-5.2',
    messages: [
      { role: 'system', content: 'You are a helpful assistant.' },
      { role: 'user', content: 'Explain quantum computing in simple terms.' }
    ],
    temperature: 0.7
  });

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

main();

Using curl#

bash
curl https://crazyrouter.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-api-key" \
  -H "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36" \
  -d '{
    "model": "gpt-5.2",
    "messages": [{"role": "user", "content": "Hello, GPT-5.2!"}],
    "temperature": 0.7
  }'

Streaming Responses#

For real-time output, enable streaming:

python
from openai import OpenAI

client = OpenAI(
    api_key="sk-your-api-key",
    base_url="https://crazyrouter.com/v1",
    default_headers={
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"
    }
)

stream = client.chat.completions.create(
    model="gpt-5.2",
    messages=[{"role": "user", "content": "Write a short story about AI."}],
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

Using Reasoning Models (o3-pro)#

The o3-pro model excels at complex reasoning tasks:

python
response = client.chat.completions.create(
    model="o3-pro",
    messages=[
        {"role": "user", "content": "Solve this step by step: If a train travels 120 miles in 2 hours, then stops for 30 minutes, then travels another 90 miles in 1.5 hours, what is the average speed for the entire journey including the stop?"}
    ]
)

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

GPT-5 Codex Models#

For code generation tasks, use the specialized codex models:

python
response = client.chat.completions.create(
    model="gpt-5-codex",
    messages=[
        {"role": "user", "content": "Write a Python function to implement binary search"}
    ]
)

Available codex variants: gpt-5-codex, gpt-5-codex-high, gpt-5-codex-medium, gpt-5-codex-low, gpt-5.2-codex

Best Practices#

  1. Choose the right model: Use gpt-5-nano for simple tasks, gpt-5.2 for complex ones
  2. Set appropriate temperature: Lower (0.1-0.3) for factual tasks, higher (0.7-1.0) for creative tasks
  3. Use streaming: For better user experience in chat applications
  4. Handle errors gracefully: Implement retry logic for rate limits

Next Steps#


For questions, contact support@crazyrouter.com

Implementation Guides

Podcast

AI Dev Tools: The Crazyrouter Podcast

Short episodes for developers on AI APIs, gateways, cost control and coding tools.

Latest episodes

  • EP196 · Oct 29, 2026 · 6:53

    AI API Security Posture — Turn Controls Into Continuous Evidence

    A practical guide to AI API security posture: inventory exposure, verify controls continuously, protect keys and data, constrain tools, and turn security assumptions into evidence that operators can act on.

  • EP195 · Oct 28, 2026 · 7:21

    AI API Policy Drift — Detect When Production Rules Stop Matching Intent

    A practical guide to policy drift in AI APIs: version routing and safety rules, detect stale enforcement, compare intended and observed behavior, and restore control before small gaps become incidents.

Episodes with written show notes →

Related Articles

Gemini 2.5 Pro and Gemini 3 Pro API Integration GuideTutorial

Gemini 2.5 Pro and Gemini 3 Pro API Integration Guide

Complete guide to integrating Google's Gemini 2.5 Pro, Gemini 2.5 Flash, and Gemini 3 Pro models via API. Includes native format and OpenAI-compatible examples.

Jan 22
WAN 2.2 Animate Tutorial 2026: Character Motion Workflows with API ExamplesTutorial

WAN 2.2 Animate Tutorial 2026: Character Motion Workflows with API Examples

A developer-focused WAN 2.2 Animate tutorial article covering what it is, alternatives, API examples, pricing, FAQs, and when to use Crazyrouter for unified routing.

Jun 6
/v1/chat/completions vs /v1/responses vs /v1/messages: Which AI API Endpoint Should You Use?Tutorial

/v1/chat/completions vs /v1/responses vs /v1/messages: Which AI API Endpoint Should You Use?

A practical guide to choosing the correct AI API endpoint. Learn the differences between OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages to avoid model unavailable errors caused by wrong endpoint routing.

Jun 4
text-embedding-3-small Dimensions Explained: 1536 vs 1024 vs 512Tutorial

text-embedding-3-small Dimensions Explained: 1536 vs 1024 vs 512

A practical guide to text-embedding-3-small dimensions: default 1536 vectors, the dimensions parameter, storage tradeoffs, and API examples.

Jun 5
AI API Gateway: Architecture, Features, and Vendor Selection GuideTutorial

AI API Gateway: Architecture, Features, and Vendor Selection Guide

Your GenAI feature can hit a wall fast: a free API tier may allow only 60 requests per minute, then return 429 errors during normal team testing. Moving to paid access may raise that to 600 request...

Mar 18
How to Get a Claude API Key in 2026: Secure Setup for Local Development and CITutorial

How to Get a Claude API Key in 2026: Secure Setup for Local Development and CI

Learn how to get a Claude API key, configure it safely, test the first request, and avoid leaking credentials in Git or CI logs.

Sep 1