I deliberately studied “API plans” rather than “API pricing,” because the price is the least interesting part. A plan defines what a class of consumer actually gets: which endpoints and operations, what rate limits and quotas, what data volumes, what support commitment, what SLA, under what terms — and then, yes, at what price.
Structure them around consumer types you can name out loud. The evaluator who wants to know if this works. The small integrator with steady modest usage. The production customer whose business depends on you. The strategic partner with a negotiated relationship. If you cannot describe a real person for each tier, you have invented a tier.
Common mistakes worth avoiding: gating the operations people need to evaluate you behind the paid tier; making the jump between tiers so large that growing consumers get punished; hiding the plan details until after signup; and letting the plans drift out of sync with what the gateway actually enforces.
Publish them in a machine-readable form alongside your rate limits. Consumers — and increasingly agents and procurement tooling — need to compare what access costs across providers, and the ones who make that easy get shortlisted by people who never spoke to sales.
