Estimate token count and API cost for your prompts across GPT-4, Claude 3.5/Opus, Gemini 1.5, and more. Helps budget AI projects before they run.
Approximately 1,333 tokens. The standard approximation is 0.75 words per token, or 1 token per ~4 characters. A typical A4 page (500 words) is roughly 667 tokens.
GPT-4o costs $5.00 per million input tokens and $15.00 per million output tokens. For a 500-token prompt with a 500-token response, each request costs about $0.01. 1,000 such requests cost approximately $10.
Claude Sonnet 4 costs $3.00 per million input tokens and $15.00 per million output tokens. Claude Opus 4 costs $15.00 input and $75.00 output per million tokens. Costs vary by model and usage.
As of 2026, Llama 3 via Groq is among the cheapest at approximately $0.05 per million input tokens. GPT-4o mini at $0.15/M and Gemini 1.5 Flash at $0.075/M are also very cost-effective for high-volume applications.
Estimate tokens per request (input + expected output), multiply by price per token for your chosen model, then multiply by monthly request volume. This calculator does all steps automatically.
A token is a chunk of text — roughly a word or part of a word. The word "hamburger" might be 1-2 tokens. "hamburgers" might be 3 tokens. Spaces, punctuation, and capitalization all affect tokenization.
Use a cheaper model for simple tasks (GPT-4o mini, Gemini Flash). Shorten system prompts. Cache common responses. Use streaming to reduce timeout retries. Consider open-source models for predictable queries.