API Usage, Rate Limits & Quotas: Token Buckets, Retries and Cost
Hands-on labs for engineers and product people who run or consume APIs: requests per second vs limits, a token bucket step by step, fixed vs sliding windows, exponential backoff with jitter, 429 rates and retry success, quotas per plan tier, batching and caching savings, and cost per customer with a margin chart, every cell checked as you go.
The first lesson is free — no account needed.
About this course
An API usage and rate limits course for engineers, product managers and technical founders who run or consume APIs, with every mechanism simulated step by step in a spreadsheet. You turn a per-minute request log into requests per second and count the minutes over the limit, simulate a token bucket with its capacity and refill rate and count the rejected requests, run the same burst through a fixed-window and a sliding-window counter, build an exponential backoff schedule with a cap and full jitter, measure the 429 rate and the success rate after retries and the extra load retries add, plan quotas, overage revenue and peak load per customer tier, work out what caching and batching save in calls and dollars and calculate cost to serve and gross margin per pricing plan with a chart. The first lab is free.
Lessons
- Requests per Second & per Minute vs the Limit: the Peak from the LogsFree
- Token Bucket Simulation: Capacity, Refill Rate, Allowed & Denied Pro
- Fixed vs Sliding Window Counters: the Boundary Burst Pro
- Retries with Exponential Backoff & Jitter: the Delay Table Pro
- 429 Error Rate & Success After Retries Pro
- Quota Planning per Customer Tier: Included Calls, Overage & Peak Load Pro
- Batching & Caching: How Many Calls (and Dollars) They Save Pro
- API Cost per Customer & Pricing Tiers: Gross Margin (Chart) Pro