Picking the right AI model for an API project often comes down to cost per token (the small chunks of text that models process and charge you for). A new free web tool makes that comparison painless by pulling pricing data for roughly 4,000 models across every major provider.
The LLM Token Cost Calculator lets you plug in your expected input and output token counts, then instantly see what each model would charge. Data comes from models.dev, an open model directory, so it stays current as providers update pricing. The interface is simple: choose a model, enter your volumes, get a number.
For anyone building apps on top of GPT-4o, Claude, Gemini, Mistral, or any of the dozens of smaller providers, this solves a real annoyance. Pricing pages are scattered, formatted differently, and sometimes buried behind login walls. Having one place to compare apples-to-apples across 4,000 models is genuinely useful, especially when you're deciding between a flagship model and a cheaper alternative that might be good enough for your use case.
The tool is free and runs entirely in the browser. It won't replace a proper cost monitoring setup for production workloads, but for quick estimates and early-stage planning, it does the job.