How does NumStack calculate AI token costs?

NumStack uses real published pricing from OpenAI, Anthropic, Google, and other AI providers. Enter your typical usage (input tokens, output tokens, requests per day) to see projected monthly costs across all models.

Why do different AI models have different token prices?

Different models have different computational costs. Smaller models like GPT-4o mini or Haiku are cheaper but less capable. Larger models like GPT-4o or Claude Sonnet are more powerful but cost more per token.

How can I reduce my AI API costs?

Use smaller models for simple tasks, cache repeated prompts, trim output length (output tokens cost 3-5x more than input), and batch requests instead of real-time when latency allows.

Contact

NumStack is a small, focused calculator platform, so email is the fastest way to reach us. Use [email protected] for anything below — mentioning the topic in your subject line helps us route it faster.

Support

Found a bug, a calculator that looks wrong, or pricing that's out of date? Email [email protected] with the calculator name and what you entered — that's usually enough for us to reproduce and fix it. You can also check the FAQ first in case it's already answered there.

Model requests

Don't see a model you use? The fastest way to request one is directly from the Token Cost calculator — scroll to the bottom of the results and click "Request it" to submit it right there. You can also email us the model name and provider and we'll add it to the queue.

Enterprise

Need higher API limits, white-label calculators, or a custom integration for your team? Email [email protected] with "Enterprise" in the subject line and a bit about your use case. See the Pricing page for what's included in each tier.

Press & partnerships

Writing about AI infrastructure costs, or interested in a data partnership or citation? Email [email protected] with "Press" in the subject line and we'll get back to you.

Response time

NumStack is run by a small team, so please allow up to 2 business days for a reply. Support and bug reports are typically the fastest to hear back on; enterprise and press inquiries may take a little longer since they often need a bit of back-and-forth.

Looking for something else?

Check the FAQ, Privacy Policy, or Terms of Service — most account, data, and billing questions are covered there.