Michael Brenndoerfer | Data & AI, Private Equity, Technology

Michael Brenndoerfer

Managing and Reducing Costs › Understanding the Cost Structure

Michael BrenndoerferAugust 26, 20252 min read

AI Agent HandbookAll chapters

Part 15 of 15Performance and Cost Optimization - Doing More with Less

Chapter 43Managing and Reducing Costs › Understanding the Cost Structure

Reading section 2 of 15 · Understanding the Cost Structure · 2 minSection 2 of 15 · 2 min

Section 2 of 15 · 2 min read

Managing and Reducing Costs › Understanding the Cost StructureOne section · Sections in this chapter2 / 15 ↓

Sections in this chapter15

Before we optimize, let's understand what you're paying for. Most language model APIs charge per token, with different rates for input and output.

Here's a simplified example of typical pricing (November 2025):

Model	Input (per 1M tokens)	Output (per 1M tokens)	Best For
Claude Sonnet 4.5	$3.00	$15.00	Complex reasoning, agents
GPT-5	$2.50	$10.00	General-purpose tasks
Gemini 2.5 Flash	$0.40	$1.20	Simple queries, high volume
Gemini 2.5 Pro	$1.25	$5.00	Multimodal, large context

Notice that output tokens cost more than input tokens. This makes sense because generating text requires more computation than processing it. It also means that verbose responses are expensive.

Let's calculate the cost of a typical interaction:

In[3]:

Code

def calculate_interaction_cost(input_tokens, output_tokens, model="claude-sonnet-4-5"):
    """Calculate the cost of a single model interaction."""
    # Pricing per million tokens (November 2025 rates)
    pricing = {
        "claude-sonnet-4-5": {"input": 3.00, "output": 15.00},
        "gpt-5": {"input": 2.50, "output": 10.00},
        "gemini-2.5-flash": {"input": 0.40, "output": 1.20},
        "gemini-2.5-pro": {"input": 1.25, "output": 5.00}
    }
    
    rates = pricing[model]
    
    # Calculate cost (rates are per million tokens)
    input_cost = (input_tokens / 1_000_000) * rates["input"]
    output_cost = (output_tokens / 1_000_000) * rates["output"]
    total_cost = input_cost + output_cost
    
    return {
        "input_cost": input_cost,
        "output_cost": output_cost,
        "total_cost": total_cost
    }

## Example: A conversation with context
input_tokens = 1500  # System prompt + conversation history + query
output_tokens = 500  # Agent's response

cost = calculate_interaction_cost(input_tokens, output_tokens, "claude-sonnet-4-5")
print(f"Input cost: ${cost['input_cost']:.6f}")
print(f"Output cost: ${cost['output_cost']:.6f}")
print(f"Total cost: ${cost['total_cost']:.6f}")
print(f"\nCost per 1000 interactions: ${cost['total_cost'] * 1000:.2f}")

Out[3]:

Console

Input cost: $0.004500
Output cost: $0.007500
Total cost: $0.012000

Cost per 1000 interactions: $12.00

Output:

Input cost: $0.004500
Output cost: $0.007500
Total cost: $0.012000

Cost per 1000 interactions: $12.00

A single interaction costs about one cent. That seems small, but multiply it by thousands of users and millions of interactions, and you're looking at serious money.