Cache Savings Simulator

This simulator compares baseline and candidate cache-hit assumptions with identical workload inputs, then estimates per-user savings, monthly savings, and break-even impact.

How It Works

1. Set provider/model plus shared usage assumptions.
2. Set baseline cache hit rate and candidate target hit rate.
3. Review savings, breakdown changes, and sensitivity ranking.

Formula

savings_per_user_month = cost_baseline - cost_candidate

monthly_savings = savings_per_user_month * monthly_active_users

Assumptions and Units

• Currency: USD
• Token unit: token
• Cache hit rates are bounded from 0 to 0.99
• Pricing source: daily snapshot in repo, no runtime scraping

Related resources: RAG Retrieval Cost Calculator, Context Window Cost Calculator, RAG Cost per User, What Cache Hit Rate Means for RAG.