GPT-4: Capabilities & Usage Guide

OpenAI's GPT-4 family includes multiple models optimized for different use cases, from fast chat to complex reasoning.

Model Variants

Model	Context	Best For
GPT-4o	128K	Multimodal, fast responses
GPT-4 Turbo	128K	Complex reasoning, long context
GPT-4o mini	128K	Cost-effective, high volume

Key Strengths

Multimodal Capabilities

GPT-4o can process:

• Text input and output
• Image analysis
• Audio (speech) input/output
• Document understanding

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What's in this image?"},
            {"type": "image_url", "image_url": {"url": image_url}}
        ]
    }]
)

Function Calling

Native support for structured tool use:

tools = [{
    "type": "function",
    "function": {
        "name": "get_weather",
        "description": "Get current weather for a location",
        "parameters": {
            "type": "object",
            "properties": {
                "location": {"type": "string"},
                "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}
            },
            "required": ["location"]
        }
    }
}]

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "What's the weather in Tokyo?"}],
    tools=tools
)

JSON Mode

Guaranteed valid JSON output:

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "List 3 programming languages as JSON"}],
    response_format={"type": "json_object"}
)

API Usage

Basic Completion

from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Explain recursion in programming."}
    ],
    temperature=0.7,
    max_tokens=500
)

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

Streaming

stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Write a Python tutorial."}],
    stream=True
)

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

Vision Analysis

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "Describe this architecture diagram"},
            {"type": "image_url", "image_url": {
                "url": "https://example.com/diagram.png"
            }}
        ]
    }]
)

Optimal Use Cases

Code Generation

Create a Python FastAPI application with:
- User authentication using JWT
- CRUD operations for a blog post model
- SQLAlchemy ORM with PostgreSQL
- Proper error handling
- Input validation with Pydantic

Data Extraction

Extract structured data from this invoice image:
- Vendor name
- Invoice number
- Line items with prices
- Total amount
- Due date

Return as JSON.

Content Creation

Write a technical blog post about microservices vs monoliths.
Target audience: Senior developers
Length: 1500 words
Include: code examples, pros/cons table, decision framework

Best Practices

System Prompts

messages = [
    {
        "role": "system",
        "content": """You are a senior software engineer.

Rules:
- Provide production-quality code
- Include error handling
- Add comments for complex logic
- Follow PEP 8 for Python
- Suggest tests for critical functions"""
    },
    {"role": "user", "content": "Build a rate limiter class."}
]

Temperature Settings

Temperature	Use Case
0	Deterministic, factual responses
0.3-0.5	Balanced creativity
0.7	Creative writing
1.0+	Highly creative, varied

Token Management

import tiktoken

encoder = tiktoken.encoding_for_model("gpt-4o")
tokens = encoder.encode("Your text here")
print(f"Token count: {len(tokens)}")

Pricing (as of 2024)

Model	Input (1M tokens)	Output (1M tokens)
GPT-4o	$2.50	$10.00
GPT-4o mini	$0.15	$0.60
GPT-4 Turbo	$10.00	$30.00

Limitations

• Knowledge cutoff date
• Token context limits
• Rate limits per tier
• Potential for hallucinations
• Cost at scale

GPT-4 vs Claude Comparison

Feature	GPT-4o	Claude Sonnet
Context	128K	200K
Multimodal	Yes (native)	Yes (images)
Function calling	Native	Via tools
JSON mode	Yes	Via prompting
Speed	Fast	Fast
Code quality	Excellent	Excellent