We use cookies for analytics and personalized ads. We use cookies for basic analytics and, where we show ads, to personalize them. Privacy PolicyRead our Privacy Policy.
We build small, free, single-purpose calculators and visualizers for learners, engineers and researchers — token and cost tools for people working with LLM APIs, plus machine learning and linear algebra tools for understanding and testing ideas. Everything runs in your browser, so you can check a token count, a cost estimate or a model result before you commit to it.
Every provider prices tokens a little differently, and those rates change more often than most teams have time to track. We got tired of doing rough mental math before every API call, so we started building the calculator we wished existed — then kept adding to it.
Since then the site has grown beyond pricing. It now also includes visualizers and calculators for machine learning and statistics — regression, PCA, neural networks, gradient descent, activation functions, confusion matrices and Bayes' theorem — and linear algebra tools such as matrix multiplication. They are made for people learning these ideas, engineers applying them, and researchers who want a quick way to explore a dataset or sanity-check a result.
Nothing here is meant to replace your provider's own billing dashboard, or established statistical and machine learning software when you publish or ship something. Think of these as the tools you reach for to explore, estimate and check your work, not the ones you reach for to reconcile an invoice or to report final results.
The text, numbers and data tables you enter, including files you open, are processed right there in your browser tab. We don't have a server in the loop, so we never receive your data and there's nothing of yours for us to store or lose.
Every pricing and memory tool starts with us actually running the free tier of the thing it estimates — an API, a local model, a hosting plan — and writing down what we found, instead of guessing at round numbers.
Model prices and context windows change often. We revisit the pricing tables regularly and mark the last-checked date on every guide so you know how fresh the numbers are.
Every calculator and visualizer on this site is free with no account and no signup wall. If we ever build something paid, the free tools stay exactly as they are.
Wherever we can, a tool shows its steps: worked examples, pencil-and-paper worksheets and ready-to-run Python and R code, so you can learn the method and check the answer instead of just copying a number.
A growing set of calculators, visualizers and guides, all reachable from the menus above.
For LLM work: token and cost estimation, local hardware memory planning, generation-speed simulation, a tokens-to-words converter, and side-by-side guides for OpenAI, Claude, Gemini, and DeepSeek pricing.
For machine learning and statistics: linear, logistic and multinomial logistic regression calculators, PCA, a trainable neural network calculator, a parameter and FLOPs counter, gradient descent and activation function visualizers, a confusion matrix calculator and a Bayes' theorem calculator.
For linear algebra: a matrix multiplication and transpose calculator that shows how every result cell is computed.
The site runs on display advertising rather than subscriptions or paywalls. That's a deliberate choice — a calculator you have to sign up for isn't one you'll reach for at 2am when you're sizing a request.
You can read exactly what that means for your data on our Privacy Policy page.
We plan and write the content on this site ourselves. We also use AI assistance in both the planning and the writing — as a tool that supports us, not a replacement for our own judgment. We tell you this because you should be able to know how what you're reading was made.
Prices come from each provider's official pricing page, which we link on every guide, and we show the date they were last checked. The machine learning and math tools use standard published methods, and many show their working and give Python and R code so you can check a result yourself.
Rimsha
Computer Science graduate and researcher in agentic AI for edge devices
Apex Tech Journal is an independent project. If something's wrong, missing, or out of date, we'd genuinely like to hear about it — hello@apextechjournal.com.