AI Pricing Calculator
Compare costs across 25+ AI models. Enter your token usage and find the cheapest option instantly.
Prices last updated: 2026-03-05. Prices may vary by provider and region. Free models available via our comparator.
How AI Model Pricing Works
Most AI providers charge per token — a unit of text roughly equal to 3/4 of a word. A 1,000-word article is about 1,300 tokens. Prices vary wildly: GPT-4 Turbo costs $30 per million output tokens, while Gemini 2.0 Flash charges just $0.40 for the same amount.
Input tokens (your prompt) and output tokens (the AI response) are usually priced differently, with output tokens costing 2-5x more. Our calculator lets you adjust the input/output ratio to match your actual usage pattern.
Free Models vs Paid Models
Several high-quality models are available completely free through services like OpenRouter. Models like Gemma 3 12B, Nemotron 30B, and Trinity Large 400B offer impressive quality at zero cost. You can test all 6 free models in our model comparator before deciding if you need a paid option.