This short lesson gives freelancers, agencies and in-house teams a number they can put in a proposal. It covers the four levers that decide what an entity audit costs and the habit of pricing every run before it starts.
What this lesson covers
- Google’s free allowance for entity analysis and Knowledge Graph lookups
- Batch processing and prompt caching for LLM passes
- Checking that caching is actually working
- Model tiering by task
- Audit costs at three site sizes
- Price it, screenshot it, then run it
Google gives away 5,000 units of entity analysis a month — roughly a thousand typical pages — and 100,000 Knowledge Graph lookups a day; the course’s whole audit used 2,917 units and cost nothing. For LLM passes, batch processing suits audit work because results are not needed instantly, and prompt caching takes most of the cost off the instruction block that repeats on every page. The lesson warns that caching is a prefix match: change one byte of the instructions and it silently misses, so the cache-read count in the response is worth checking.
Model tiering is usually the largest saving of all: high-volume extraction suits the smallest model, analysis suits a mid-tier model, and the most capable model is kept for low-volume rewriting and briefs. Put together, a 1,000-page audit costs nothing, 10,000 pages about $46 and 50,000 pages about $229, and adding an LLM pass costs tens of dollars rather than hundreds when batching and caching are used. The lesson closes with the habit of checking the estimate, screenshotting it and keeping it with the audit.
The figures are list prices checked against Google’s pricing page in September 2026, and the lesson recommends re-checking them and stating the date whenever a quote goes to a client. The three site sizes are derived from the demo corpus’s measured average of 4.58 units per page, so they should be described as estimates at that page length; your own corpus will differ, and the arithmetic is reproducible either way.
This lesson completes Module 2, Diagnostics: Extracting the Truth About Your Site, in The AI-powered Entity Audit, and leads into Module 3 on the Knowledge Graph. It suits freelancers and agency teams who price audits for clients, and in-house SEOs who need budget approval before running API-based analysis.