AI API token and cost calculator
If you are budgeting a product that calls GPT, Claude, Gemini or similar APIs, this page helps you see—before you buy credits or go live—rough USD cost per request, the bill for 1,000 similar calls, and a monthly projection from how often you expect to call the API. Paste real prompts and sample assistant replies to improve token estimates, or switch models to compare scenarios. Nothing you enter is sent to any provider; bundled prices are starting points only—always confirm current tariffs on each vendor’s official site.
Pick a model, paste or type a prompt, and adjust token counts. We estimate input tokens from text length (~4 characters per token)—rough for English-like text, not a real tokenizer. Optional response text drives an output-token estimate unless you override the number.
When you change the model, bundled default rates load for that model. Edit the price fields to match your provider’s current tariffs. This tool does not call any API.
Choose model
Prompt & token usage
0 characters · ~0 tokens (estimate)
Updates from the prompt estimate when you edit the prompt unless you override this field.
If you add text here, we estimate output tokens the same way as for the prompt (~4 characters per token). Leave empty when you only care about input cost.
0 characters · ~0 tokens (output estimate)
Updates from the response text when you edit it unless you override this field. Use 0 if you expect no completion.
Used to project monthly spend: cost per request × this value.
API price per 1M tokens (optional)
Values are filled from the built-in catalog when you select a model. Adjust them as needed, or reset to bundled defaults for the current model.