Free browser tool
Plan the cost of an AI agent across its full workflow
Compare token, agent, ROI and support scenarios with practical cost guides and budget templates. Use your own rates and volumes to trace each assumption.
Tasks, steps, retries and overhead
Use tasks per month and expected model/tool steps per task. Input and output rates are per one million tokens. Cached-input fraction and cached-rate ratio range from zero to one. Extra-attempts fraction is additional attempts divided by base attempts: 0.1 means 10% extra work. Include tool cost per step, fixed monthly costs and human monthly costs.
All rates and costs must use the same currency. Example amounts are fictional assumptions. No vendor rates are preselected and no API calls are made.
Use measured assumptions
An agent can make several model and tool calls per task. Cost estimates should include context, retries, external tools, fixed costs and human review. The main calculator exposes those inputs and provides sensitivity scenarios.
Open the agent cost calculator
Keep cost and outcomes together
Use the API and token calculators for measured usage, the ROI calculator for time-value and payback assumptions, and the support calculator for escalation scenarios. No result promises cash savings or vendor performance. Everything runs locally without paid API calls.
Calculators, budget template and guides
Choose the right calculator for your question
Start with the AI agent cost calculator for tasks, model steps, tokens, retries, tools, and monthly operating costs. Use the LLM API cost calculator to isolate input, output, cached-token, and batch assumptions. The token cost calculator needs measured token counts from the model or provider. Use the ROI calculator to compare entered time-value and setup assumptions, and the support calculator to model escalations against a human-only scenario.
Each calculator works with numbers you enter in this browser. None fetches live rates, sends prompts to a model, or verifies an invoice. Use the current official rate page or your contract, keep a date and currency with each assumption, and treat outputs as scenarios to review. The guides and budget template linked below explain cost categories, scope, pricing structures, FinOps practices, and how to keep direct charges separate from allocated staff time.
Start with the question you need to answer
Estimate one agent workflow at a time, and keep the measured usage period, currency, contract and operational scope with every assumption. Use the agent calculator for steps, retries, tool calls, fixed overhead and review effort. Use the API calculator for input, output, cache and batch-rate assumptions, and the token calculator when you already have token counts. The ROI calculator separates time value from cost, while the support comparison models escalations.
Guides cover ownership cost, project build estimates, pricing structures and FinOps; the budget template provides a copyable planning worksheet. Every calculator uses the rates you enter, runs in the browser and does not fetch live prices or make model calls. Treat outputs as scenarios, compare them with invoices and operational evidence, and review them when usage or service terms change.
Updated 2026-10-08. Sources are linked on this page.
Primary sources and review
- OpenAI: official API pricing and cost units
- Anthropic: official pricing, caching and batch rates
- NIST AI RMF: risk-management work to account for in operating costs
Published by AI Agent Cost Calculator. Last updated: . Methods on this site are practical workflows; outputs do not certify compliance.