AI Economics practical guide
How to Calculate AI Cost: Why Token Price Is Not Enough
Token price is only one layer. Total AI cost also includes staff effort, cloud, RAG, agent tool calls, vector databases, observability, SaaS and implementation support. Organisations should measure the total cost of each completed workflow.
Five layers of total cost
List cloud infrastructure, AI APIs, private AI, application architecture, and SaaS or AI seats, then attribute each layer to teams, projects and workflows. Token price alone misses agent loops, OCR, retrieval, storage, network and support effort.
Use cost per completed workflow
Choose a useful business unit such as a completed invoice, reviewed enquiry or approved report. Put staff time, AI and cloud cost, and amortised implementation cost over the same denominator to compare before and after.
Management checklist
- Can AI cost be mapped to a team, project or workflow?
- Are model routing, caching, batching and execution limits in place?
- Are speed, quality, review rate and output measured together?
- Are pricing currency, units and review date clear?
Method references and review date
These sources provide related definitions, frameworks or implementation considerations. The method is synthesised by iGears and is not a fixed provider quotation or an outcome promise.
Last reviewed: 30 August 2026
Book an AI consultation
Confirm usable AI routes, implementation scope and ongoing care around your tools, data requirements and target work.