Fusion AI

Documentation

What is Fusion AI?Key Benefits How NeuroSwitch Works Data Flow

Test Your First Call

Make your first API request to Fusion AI and see NeuroSwitch intelligent routing in action. Get up and running in under 2 minutes.

Prerequisites

You have your Fusion AI API key from the quickstart guide

Option 1: cURL (Quick Test)

Terminal

curl -X POST https://api.mcp4.ai/chat \
  -H "Authorization: Bearer sk-fusion-your-api-key-here" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Explain quantum computing in simple terms",
    "provider": "neuroswitch",
    "max_tokens": 150
  }'

Important: Replace sk-fusion-your-api-key-here with your actual API key.

Expected Response

{
  "response": "Quantum computing is like having a super-powered calculator that can explore many solutions simultaneously...",
  "provider_used": "claude-3-opus",
  "routing_reason": "Educational content - Claude excels at clear explanations",
  "tokens_used": 142,
  "cost": 0.00213,
  "request_id": "req_abc123",
  "timestamp": "2024-01-15T10:30:00Z"
}

NeuroSwitch Decision

See exactly why this model was chosen

Real-time Costs

Transparent token and cost tracking

Provider Info

Know which model handled your request

Option 2: JavaScript/Node.js

JavaScript

const response = await fetch('https://api.mcp4.ai/chat', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer sk-fusion-your-api-key-here',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    prompt: 'Write a Python function to calculate fibonacci numbers',
    provider: 'neuroswitch',
    max_tokens: 200
  })
});

const data = await response.json();
console.log('AI Response:', data.response);
console.log('Model Used:', data.provider_used);
console.log('Cost:', data.cost);

Option 3: Python

Python

import requests

response = requests.post(
    'https://api.mcp4.ai/chat',
    headers={
        'Authorization': 'Bearer sk-fusion-your-api-key-here',
        'Content-Type': 'application/json'
    },
    json={
        'prompt': 'Analyze this data and provide insights',
        'provider': 'neuroswitch',
        'max_tokens': 300
    }
)

data = response.json()
print(f"AI Response: {data['response']}")
print(f"Model Used: {data['provider_used']}")
print(f"Tokens Used: {data['tokens_used']}")
print(f"Cost: $\{data['cost']}")

Try Different Providers

Test how different providers handle the same prompt. Change the provider parameter:

Let NeuroSwitch Choose

"provider": "neuroswitch"

Automatically selects the best model for your prompt

Specific Providers

"provider": "openai"

"provider": "claude"

"provider": "gemini"

Common Parameters

Required Parameters

prompt

Your message or question to the AI

provider

Which AI provider to use (or "neuroswitch")

Optional Parameters

max_tokens

Maximum response length (default: 1000)

temperature

Response creativity (0.0-1.0, default: 0.7)

stream

Enable streaming responses (true/false)

Common Issues

❌ "Unauthorized" Error

Check that your API key is correct and includes the sk-fusion- prefix.

❌ "Invalid JSON" Error

Ensure your request body is valid JSON and includes the Content-Type: application/json header.

❌ Slow Response

First requests may take 2-3 seconds as models "warm up." Subsequent requests are typically sub-second.

Great! What's Next?

You've successfully made your first API call. Now explore more advanced features and integrations.

Set Up Chat Sessions

Learn how to maintain conversation context

Enable Streaming

Get real-time token-by-token responses

Use SDKs

Integrate with official libraries and tools