Claude Code Pricing: What Each Plan Includes and What It Costs
A dated explainer: which plans include Claude Code, how the usage limits work, what the docs say it costs per developer, and the habits that spend less.
Claude Code pricing comes down to one choice: a subscription plan, where usage is included up to a limit, or pay-per-token through the API. On Anthropic's pricing page the lowest-priced subscription that includes Claude Code is Pro, at $20 a month or $17 a month billed annually. Plan prices and documented limits below come from two public pages, fetched on 2026-10-26; the "What I pay" section describes my own subscriptions. Short version: Pro to try it, Max when you hit limits regularly, Team or Enterprise for organisations, and the API route when you want to pay per token. Prices change, so check the live pages before you buy.
Updated 2026-10-26. Sources: claude.com/pricing and code.claude.com/docs/en/costs, both fetched that day.
Which plan includes Claude Code
The pricing page's FAQ says: "Claude Code is included in all paid plans." The Free card does not list Claude Code. Here is each plan as the page showed it on 2026-10-26. The usage column is the page's own wording, not mine:
| Plan | Price on the page | Claude Code | Usage statement on the page |
|---|---|---|---|
| Free | $0 | Not listed | "Pro gives you at least 5x more usage per 5-hour session than Free." |
| Pro | $20 monthly, or $17 a month with annual billing ($200 billed up front) | Included | "At least 5x more usage per 5-hour session than Free" (the page's Pro vs. Free line) |
| Max | From $100 a month | Included | "Max gives you 5x or 20x more usage per 5-hour session than Pro." |
| Team, Standard seat | $25 monthly, or $20 a month annually | Included in paid plans | Seat-based, for teams of 2 to 150; no multiplier stated |
| Team, Premium seat | $125 monthly, or $100 a month annually | Included in paid plans | "5x more usage than standard seats" |
| Enterprise | US$20 per seat per month, billed annually, 20 seat minimum, plus usage at API rates | Included in paid plans | "Usage cost scales with model and task." |
The table's point: only Enterprise adds usage billed at API rates on top of the seat price; Pro, Max and Team include limited usage in their subscription prices, and usage credits cost extra. If I had to pick from the table alone: Pro to try it on real work, Max when you hit the Pro limit regularly, Team or Enterprise when an organisation needs admin controls. The page lists spend limits, audit logs and custom data retention for Enterprise. The page does not say how many messages or tokens a "5x" session holds, and neither do I.
How the usage limits work
Per the pricing FAQ: "Every plan has usage limits that reset on a rolling five-hour session window, and paid plans add weekly limits on top." The FAQ also says Claude Code "shares the same usage limits as the rest of your plan", so chat on the same account draws from the same allowance.
The costs documentation adds the practical details:
- Session and weekly limit messages are shared across models. After "You've hit your session limit" or "You've hit your weekly limit", switching models with
/modeldoes not restore access. The message shows when the window resets. - Model-specific messages are different. After "You've hit your Opus limit" or "You've hit your Sonnet limit", switching to a model outside that family does keep you working.
- Usage credits extend a plan. Per the docs, usage credits "let you keep working past your plan's usage limit", managed with
/usage-creditsafter signing in with a claude.ai subscription. They are not available with API key authentication. - Waiting is an option. On Claude Code v2.1.234 or later, the docs say Claude Code can wait for the reset and continue the interrupted task automatically.
To see where you stand, /usage shows plan usage bars on a subscription. The same page notes that the dollar figure in its Session block is "intended for API users" and is computed locally at list price by default (an administrator can configure other rates), so on Pro or Max it is not what you are billed.
The pay-per-token route
The docs open with the other model: "Claude Code charges by API token consumption." If you authenticate through the Claude Console or a cloud provider instead of a subscription, usage is billed per token to your organisation. The docs give one benchmark: across enterprise deployments, the average cost is "around $13 per developer per active day and $150-250 per developer per month, with costs remaining below $30 per active day for 90% of users."
That is an enterprise average, not a quote for your project. The docs themselves say to "start with a small pilot group" and measure a baseline, and that per-developer costs vary with model choice, codebase size and usage patterns such as running multiple instances. I am not reproducing per-model token prices here; the live API pricing page is the place for those.
What I pay
I run two Claude plans at $200 a month each, plus one Team seat. The pricing page lists Max as "From $100" with a choice of 5x or 20x more usage than Pro, and the text I fetched does not print a $200 subscription price, so I am placing mine in the Max tier and not claiming a multiplier. The Team seat has lower limits than my $200 plans; I do not know its exact limits, so I give none.
What I did not measure
I have not logged my own Claude Code usage against plan limits, so this post has no "here is how fast I hit Pro" number. Any figure like that would be a guess, and the docs' own warning applies: usage varies widely. If you want your own number, the /usage plan bars and the breakdown beneath them are the tool, and the docs say the breakdown recommends a change for behaviours behind 10% or more of your recent usage.
How to spend less, per the docs
The costs page lists the levers. The ones I find most useful, each from that page:
- Clear between tasks.
/clearwhen switching to unrelated work, because "stale context wastes tokens on every subsequent message." - Match the model to the job. The docs say Sonnet "handles most coding tasks well and costs less than Opus", and to reserve Opus for complex architectural decisions or multi-step reasoning.
- Lower the effort level for simpler tasks, with
/effort. - Delegate verbose work to subagents, so test output stays out of your main context. The docs add that a subagent's own requests still draw on your usage. I wrote about when that is worth it in Claude Code Subagents: When One Beats Doing It Inline.
- Keep CLAUDE.md short. It loads at the start of every session, and the docs suggest moving workflow-specific instructions into skills and keeping the file under 200 lines. My own file is the cautionary example: CLAUDE.md File: What to Put In It and What to Leave Out has the size history.
- Watch long sessions. The docs list long context, cache misses after a break, scheduled tasks and subagents as reasons a long session uses more of your limits than your activity suggests.
What this page does not cover
No comparison with other tools' prices, no claim about what Anthropic will charge later, and no per-model API price table. This post is a snapshot of two pages as they read on 2026-10-26. If the numbers above no longer match the live pages, trust the live pages.
Keep reading
- Gemini CLI vs Claude Code: What I Can and Cannot Say From My LogsA correction first: none of my logged runs used Gemini CLI. Here is what the Gemini rows are, what the two tools' docs say, and a dated tally with its limits.
- Using Codex as the Read-Only Reviewer for Claude CodeNot a benchmark: Claude Code writes and orchestrates, Codex reviews. The split, the verdict contract, and one case where the gate said no.
- CLAUDE.md File: What to Put In It and What to Leave OutWhat a CLAUDE.md file costs, what the docs say belongs in it, and the size history of mine: 3,639 words, then 512, then 9,555.