From the Free plan to Enterprise, plus the API usage tiers behind every automation — here's how Claude's pricing and limits actually work, and how you raise a ceiling like the API's per-minute token limit.
Two pricing worlds: subscriptions and the API
Claude has two separate pricing models, and most businesses use both. Subscriptions cover people using the Claude app day to day; the API (pay-per-token) powers the automations, tools, and products you build on top of it. Knowing which is which keeps your budgeting sane.
The subscription tiers
There's a Free tier to try it, Pro (around $20/month) as the mainstream professional plan with the latest Opus model, Claude Code, and Cowork, and Max (roughly class="fs-unmask-airo-app-builder"00–$200/month) for heavy daily users who need several times the capacity. For organizations, Team is priced per seat with admin controls and shared Projects, and Enterprise is custom with SSO, data controls, and compliance. Pricing changes, so confirm current numbers at claude.com/pricing.
How app usage limits work
Usage in the Claude app is on a rolling limit that replenishes every few hours rather than a hard monthly cap, and higher plans raise the ceiling. The practical rule: if you're regularly bumping into limits during real work, that's your signal to move up a tier — the cost of the upgrade is usually trivial against the time lost waiting.
The API usage tiers, where automations live
API access is metered per token and organized into usage tiers. Each tier sets per-minute ceilings — requests per minute, input tokens per minute, and output tokens per minute — and new accounts start at Tier 1. These per-minute throughput limits, not your monthly budget, are usually what you hit first when you scale up an automation.
Raising a limit — the "30k" question
Tier 1 historically started around 30,000 input tokens per minute, so a high-volume job could exhaust a minute's budget in one or two large requests and start queuing. You raise the ceiling by advancing usage tiers (tied to cumulative spend), requesting a higher or custom limit in the Anthropic Console, or committing to Priority Tier for guaranteed capacity. Anthropic has also increased these limits substantially over time — so if you abandoned a workflow because of rate limits a while ago, it's worth retesting.
Scale smarter, not just bigger
Before you buy more capacity, architect better. Route simple tasks to Haiku (cheapest, most generous limits) and reserve Opus for reasoning that truly needs it, and use prompt caching so repeated context doesn't burn your token budget on every call. Smart design often beats a bigger plan — and that's usually where the expertise pays for itself.
Need Help?
Sizing plans, managing API costs, and scaling limits is exactly the kind of thing we help clients get right. PreciseHR offers a free 30-minute consult.
Explore PreciseHR tools
About the author
Jeffrey T. Furtado
Managing Partner, PreciseHR
Jeffrey T. Furtado (Jeff Furtado) is an executive leader, entrepreneur, and investor with a track record of building, scaling, and transforming businesses. As both a corporate operator and founder, he has led high-growth teams, driven operational excellence, and helped create lasting enterprise value. He writes about leadership, execution, strategy, and building organizations that stand the test of time.
More about Jeffrey