{"head":{"title":"AI Agent Cost Calculator: Cost per Task | OceanLab","description":"Estimate what an AI agent costs per task and per month from its steps, tool calls and tokens, across models. Includes context growth on every step.","canonical":"https://oceanlab.in/tools/ai-agent-cost"},"tool":{"slug":"ai-agent-cost","path":"/tools/ai-agent-cost","name":"AI agent cost","h1":"AI agent cost calculator","lede":"Estimate what an AI agent costs per task and per month, including the context it re-sends on every step.","category":"ai","categoryTitle":"AI cost and planning","method":["An agent re-sends its whole history on every step. Step i sends the base prompt (system prompt, tool definitions and task) plus everything produced in steps 1 to i−1: model output and tool results.","The base prompt is the part a prompt cache can serve; the history is uncached by default. Retries add a share of extra steps.","The tokens for one task are priced on every model at its published rates, cheapest first, then multiplied by tasks per month."],"sources":[{"label":"Anthropic API pricing","url":"https://platform.claude.com/docs/en/about-claude/pricing","checkedAt":"2026-10-23"},{"label":"OpenAI API pricing","url":"https://developers.openai.com/api/docs/pricing","checkedAt":"2026-10-23"},{"label":"Google API pricing","url":"https://ai.google.dev/gemini-api/docs/pricing","checkedAt":"2026-10-23"},{"label":"DeepSeek API pricing","url":"https://api-docs.deepseek.com/quick_start/pricing","checkedAt":"2026-10-23"},{"label":"MiniMax API pricing","url":"https://platform.minimax.io/docs/pricing/overview","checkedAt":"2026-10-23"}],"faq":[{"question":"Why do agents cost more than single calls?","answer":"An agent re-sends its history on every step, so each step's input includes all earlier outputs and tool results. Cost grows faster than the number of steps."},{"question":"What counts as a step?","answer":"One model call. A step may request one or more tool calls, and their results are added to the next step's input."},{"question":"Does caching help?","answer":"Yes. The system prompt and tool definitions repeat on every step and can be served from the provider's prompt cache at a lower rate."},{"question":"Where do the prices come from?","answer":"Each price is read from the provider's official pricing page and shows the date it was checked. Prices that could not be verified are left out, or start at 0 for you to fill in."}],"updatedAt":"2026-10-25"},"related":[{"href":"/tools/llm-api-cost-calculator","label":"LLM API cost","detail":"Enter your monthly tokens and the input, cached and output mix, and compare what each model would cost."},{"href":"/tools/prompt-caching-savings","label":"Prompt caching savings","detail":"See how much prompt caching would save on each model, and the hit rate you need before it pays off."},{"href":"/tools/ai-automation-roi","label":"AI automation ROI","detail":"Put in one task your team does by hand and see the monthly saving, the payback period and the net result over three years."}],"studies":[{"meta":{"slug":"sanbuk","industry":"ai","projectType":"Back-office agent","country":"ae","tags":["ai","enterprise"],"compliances":["UAE VAT","FTA tax invoice rules","UAE e-invoicing (Peppol)","UAE Corporate Tax records","UAE PDPL","UAE and UN sanctions lists","OWASP Top 10 for LLM Applications","UAE AI Charter","ISO/IEC 42001"],"cover":{"src":"/case-studies/sanbuk/01-home.png","width":1440,"height":1000}},"text":{"en":{"title":"Sanbuk - Accounts Payable Agent with an Operator Console and Approval Gates","description":"Designed and built a clickable back-office agent for a Dubai mainland distributor: it reads the AP and vendor mailboxes, matches invoices to POs and goods receipts, checks UAE tax invoice fields, drafts replies and posts to the ERP, while people approve by amount, see every tool call and handle every bank detail change themselves.","category":"AI Agents & Automation / Back-Office Agent","client":"Ghafline Distribution LLC"}}},{"meta":{"slug":"anvesha","industry":"ai","projectType":"Research agent","country":"in","tags":["ai","fintech","enterprise"],"compliances":["SEBI RA Regulations","SEBI PIT Regulations (UPSI)","SEBI LODR","Copyright Act 1957","RFC 9309 (robots.txt)","DPDP Act 2023"],"cover":{"src":"/case-studies/anvesha/01-home.png","width":1440,"height":1000}},"text":{"en":{"title":"Anvesha - Deep-Research Agent for Cited Equity Research","description":"Designed and built a clickable deep-research agent for a SEBI-registered Mumbai research boutique: analysts brief the agent, watch each run stay inside its time, cost and page budget, and get a draft report in which every factual sentence links to a quoted source passage they must accept before it can leave the team.","category":"AI Agents / Research Agent","client":"Mahim Bay Research"}}},{"meta":{"slug":"lektor","industry":"ai","projectType":"Code review agent","country":"ch","tags":["ai","enterprise","fintech"],"compliances":["Swiss Banking Act Art. 47","FINMA Circular 2023/1","FINMA Guidance 08/2024","revFADP","ArGV 3 Art. 26","OWASP Top 10","CWE"],"cover":{"src":"/case-studies/lektor/01-home.png","width":1440,"height":1000}},"text":{"en":{"title":"Lektor - AI Code Review Agent That Comments but Never Merges, Hosted Inside a Private Bank","description":"Designed and built a clickable AI code review agent for the engineering team of a Zurich private bank: it reviews every merge request on GitLab and GitHub Enterprise with models hosted in the bank's own tenancy, masks client data and secrets before any model sees the diff, posts inline findings with rule, CWE and suggested patch, and never approves or merges.","category":"AI Agents & Automation / Code Review Agent","client":"Brunnacher & Cie AG"}}},{"meta":{"slug":"kaito","industry":"ai","projectType":"Meeting notes to CRM","country":"hk","tags":["ai","enterprise"],"compliances":["PDPO (Cap. 486)","PCPD AI Model Framework","Estate Agents Ordinance (Cap. 511)","AMLO (Cap. 615)"],"cover":{"src":"/case-studies/kaito/01-home.png","width":1440,"height":1000}},"text":{"en":{"title":"Kaito - Meeting Notes to CRM Assistant with Field-Level Review","description":"Designed and built a clickable meeting assistant for a Hong Kong commercial property brokerage: it transcribes brokers' meetings and calls in Cantonese, English and Mandarin, proposes CRM changes as a field-level diff with evidence from the transcript, flags likely duplicate contacts, and writes nothing to Salesforce until the agent approves.","category":"AI Agents & Automation / Meeting Notes to CRM","client":"Stonecutter Lane Property Partners"}}}],"more":[{"href":"/case-studies/industry/ai","label":"AI agents & automation case studies","detail":"Every OceanLab product build in AI agents & automation."},{"href":"/case-studies/industry/internal","label":"Internal tools & ops case studies","detail":"Every OceanLab product build in Internal tools & ops."},{"href":"/case-studies","label":"All case studies","detail":"Every OceanLab product build."},{"href":"/ai-automation","label":"AI automation","detail":"AI built into real business workflows, with people in charge."}],"today":"2026-10-25"}