AI & LLM Cost Calculators
2 free tools · no login · nothing uploaded
There are two quite different AI cost questions and most calculators only answer the first one.
The first is what a request costs: input tokens times a rate, output tokens times a rate, adjusted for prompt caching and batch discounts. That is the right model for a chat feature or a classification endpoint, and the token calculator here does it across more than thirty models — including the detail almost everyone skips, which is that some providers charge more to WRITE to the cache than to read normal input, making caching a loss below a break-even hit rate.
The second question is what an agent costs, and it has a different shape entirely. An agent does not make a call; it runs a loop, and every step re-sends the accumulated transcript. Step twelve pays again, as input, for everything the first eleven steps read and produced. Summed across a run that grows with the square of the step count, not linearly — so doubling an agent's steps roughly quadruples its cost. Add sub-agent fan-out and retries and a single task can cost tens of times what a token calculator would quote for the same prompt.
That amplification is why AI budgets get blown by loop length rather than by rate, and it is what the agent calculator models.
2 tools in ai & llm cost
LLM Token Cost Calculator
Model your AI spend across 30+ models before the invoice arrives.
Compare what a real workload costs on Claude, GPT, Gemini, Nova, and open models — with prompt caching, batch discounts, and per-conversation unit economics. Includes an agentic budget-overrun simulator.
AI Agent Cost Calculator
Agents do not make one call. Price the loop, not the request.
Model what an agent actually costs: steps per run, sub-agent fan-out, context that grows as the transcript accumulates, cache hits, and retries. Gives cost per run, per developer and per month — and the date your budget runs out.
Other categories
A model is a forecast. Your invoice is the measurement.
Finitizer tracks real token consumption per agent, per feature and per team across Anthropic, OpenAI, Bedrock and Vertex — so you learn that a loop got longer from a trend line, not from a budget alert on the 22nd.