ToolPuma Logo

AI Model Comparison

Compare context windows and pricing across providers.

ModelProviderContextMax outputInput / 1MOutput / 1MCost
GPT-4.5 Previewopenai128,00016,384$75.00$150.00$$$
o1openai200,000100,000$15.00$60.00$$$
o3-miniopenai200,000100,000$1.10$4.40$$
GPT-4oopenai128,00016,384$2.50$10.00$$
GPT-4o miniopenai128,00016,384$0.150$0.600$
GPT-4 Turboopenai128,0004,096$10.00$30.00$$$
GPT-4openai8,1924,096$30.00$60.00$$$
GPT-3.5 Turboopenai16,3854,096$0.500$1.50$$
Claude 3.7 Sonnetanthropic200,00064,000$3.00$15.00$$$
Claude 3.5 Sonnetanthropic200,0008,192$3.00$15.00$$$
Claude 3.5 Haikuanthropic200,0008,192$1.00$5.00$$
Claude 3 Opusanthropic200,0004,096$15.00$75.00$$$
Claude 3 Haikuanthropic200,0004,096$0.250$1.25$
Gemini 2.0 Pro (Exp)google2,000,0008,192$1.25$5.00$$
Gemini 2.0 Flashgoogle1,000,0008,192$0.100$0.400$
Gemini 1.5 Progoogle2,000,0008,192$1.25$5.00$$
Gemini 1.5 Flashgoogle1,000,0008,192$0.075$0.300$

Prices per 1M tokens. Context and max output are in tokens. Provider pricing changes frequently — confirm current rates before budgeting.

What Is a AI Model Comparison?

The AI Model Comparison tool lists large language models from OpenAI, Anthropic Claude, and Google Gemini side by side, showing context window, maximum output, input and output prices per million tokens, and a relative cost indicator.

How to Use This AI Model Comparison

  1. 1
    Scan the tableCompare context windows and max output across providers.
  2. 2
    Compare pricesCheck input and output rates per million tokens.
  3. 3
    Use the cost indicatorThe dollar badges give a quick relative cost sense.

Frequently Asked Questions

Consider the context window needed for your task, the max output for long generations, cost for your request volume, and quality for the specific use case.
They reflect published rates at the time of writing. Providers update pricing regularly, so confirm current figures before making decisions.
It compares window sizes and pricing, not reasoning quality. Benchmarks and your own testing should inform capability comparisons.