Tech Meridian← MODEL PASSPORTS

DEVELOPER TOOLS · FREE

What will your AI cost?

Set your workload and compare API costs for up to four models. Estimates update instantly, with no account needed.

1. Your workload
One request means one API call, not a whole conversation. Average, including instructions, history and documents. Up to 100,000. Average, including billable reasoning tokens. Up to 16,000.
Where do I find token counts?

Use averages from your API response usage fields or provider usage reports. Tokens are pieces of text, not words. Models tokenize text differently; this estimate uses the same token counts for each model.

2. Compare models

Same workload · standard text API · no caching

30-day estimate

This compares cost, not answer quality. Prices are in USD, excluding taxes and additional services.

How we calculate and which rates we use

Monthly cost = 30 × daily requests × (input tokens × input rate + output tokens × output rate) / 1,000,000. Workload presets are starting examples, not measurements of your product.

We use the provider's direct paid API: text, standard processing, uncached input. Free allowances, batch processing, discounts, faster processing, regional surcharges, tools, search, images and audio are excluded. Include billable reasoning in output tokens. This calculator supports up to 100,000 input and 16,000 output tokens per request; it does not estimate long-context workloads with different rates.

Rates checked against official sources: . Check the provider's terms before launching: prices and availability can change.

Explore model capabilities and limitations →