Start with a real workload, then see the cheapest sensible model, the expensive outliers and the trade-off in plain money. No sign-up, no pasted text sent anywhere.
No sign-up · runs entirely in your browser
Type a word count, paste any text, or pick a content type. Drag the conversation slider. Toggle Batch and Cache. Everything you just saw — but live, with your numbers.
—
It depends on the model and how much text goes in and out. TokenScale shows real examples. Writing the whole of The Hobbit costs about $0.06 on Gemini Flash-Lite, one of the cheapest tiers tracked. The same job costs far more on a frontier model. Pick a content size and a provider to see the live figure.
Most providers charge more for the tokens a model generates than for the tokens you send. Output often costs three to five times more than input. A long answer to a short question can cost more than it looks. TokenScale splits every price into input and output so the gap is visible.
A lot. Across the 22 providers TokenScale tracks, input rates span roughly 500× — from $0.02 to $10 per million tokens — and even models aimed at similar work are routinely more than thirty times apart. Choosing the right provider for a task can cut the bill by an order of magnitude. The comparison table sorts every provider cheapest first.
Three things lower the bill. Pick a low-cost provider, use Batch mode where it is offered for fifty percent off, and reuse cached context for up to ninety percent off input. TokenScale lets you toggle Standard, Batch and Cache to see each saving. Open-weight models on inference hosts are usually the cheapest of all.
Yes. TokenScale is free, with no sign-up and no account. It runs entirely in your browser. Nothing you type is sent anywhere or stored.
Every night. An automated check reads each provider's pricing and records it, building a verified price history you can scroll back through. The latest verified date is shown on the site.