Which Claude plans include Claude Code? Every paid one, and the limits

The plans, as of October 9, 2026

• Free ($0): no Claude Code. Anthropic’s setup page says “The free claude.ai plan does not include Claude Code access.”
• Pro ($20 a month, or $200 a year up front, which works out to $17 a month): Claude Code with Opus and Sonnet, which Anthropic’s product page calls “perfect for short coding sprints in small codebases.” Fable, Anthropic’s top model, is reachable only through usage credits.
• Max 5x ($100 a month) and Max 20x ($200 a month): five or twenty times Pro’s usage per five-hour session, monthly billing only, with Fable included (how it counts against the week is one of the points below where Anthropic’s pages differ). The product page calls 5x “great value for everyday use in larger codebases” and 20x the plan for “power users.”
• Team, Standard seat ($25 a month, $20 billed annually): 1.25 times Pro’s usage per session, for 2 to 150 members. Fable needs usage credits.
• Team, Premium seat ($125 a month, $100 billed annually): 6.25 times Pro’s usage per session, Fable included. Team limits are per member, and seats can be mixed.
• Enterprise ($20 a seat a month, billed annually): Claude Code comes with the single Enterprise seat, but the seat fee “doesn’t include any usage. All usage across Claude, Claude Code, and Cowork is billed separately at standard API rates,” with no per-seat limits. Self-serve starts at 20 seats, and sales-assisted plans at 50. Older Enterprise contracts include Claude Code only on certain seat types.

The figures are from Anthropic’s pricing page and its help articles on Max, Team and Enterprise. Anthropic says it does not offer standing discounts on Pro or Max. On every paid plan Claude Code’s default model is Opus 5.5; before version 2.1.280 Pro and Team Standard defaulted to Sonnet 5.

The plans also differ in what Claude Code can do beyond the terminal. Per Anthropic’s feature table: cloud sessions, routines and the Claude app’s Code tab come with every paid plan; Remote Control, which lets you steer a session on your own machine from your phone, is on for Pro and Max and needs an admin to switch it on for Team and Enterprise; computer use and Dispatch are Pro and Max only; and Code Review, SSO and managed settings are Team and Enterprise only.

How the limits work

The pricing page puts it in one paragraph: “Every plan has usage limits that reset on a rolling five-hour session window, and paid plans add weekly limits on top. Your activity across Claude on web, desktop, mobile, and Claude Code all draws from the same pool.” What a window holds depends on the length of the conversation, the model, the features and, per the limits article, “the effort level you’ve selected,” so “there’s no fixed message count.” The weekly limit resets at “a fixed time each week that is assigned to your account,” whenever you subscribed; Settings, then Usage, in the Claude app shows it, and inside Claude Code /usage shows your plan’s bars and reset times (Anthropic’s Pro and Max article names /status).

The multiples are the only sizes Anthropic publishes, and all of them are per five-hour session: Pro at least five times Free, Max five or twenty times Pro, a Team Standard seat 1.25 times Pro and a Premium seat 6.25 times. How the weekly limits compare between plans is not stated anywhere. Every plan page also keeps the right to limit usage “in other ways, such as weekly and monthly caps or model and feature usage, at our discretion,” and the Claude Code terms add that “advertised usage limits for Pro and Max plans assume ordinary, individual usage of Claude Code and the Agent SDK.”

Two things matter more in Claude Code than in chat. Both limits count at once, so, in the words of the error reference, “a single burst of heavy activity, such as a large workflow fanout, can exhaust the weekly allowance before the session window resets”: many subagents or agents at once spend the week, not just the window. And when a limit hits, Claude Code says which one: “You’ve hit your session limit,” “your weekly limit,” or, the same page lists, an Opus or a Sonnet limit, which apply only to that model family, so switching models with /model keeps you working where it would not for the other two.

When you hit a limit

• Wait. In an interactive session signed in with a plan, Claude Code “waits in the open session and continues the task on its own after the limit resets” (version 2.1.234 and later).
• Usage credits. On Pro, Max and Team you can turn on usage credits: you prepay, Claude carries on past the limit at standard API rates, charged separately from the subscription, until your balance or a monthly spend limit you set runs out. Anthropic calls the same thing “extra usage” on some pages.
• A free reset. Anthropic occasionally gives eligible plans a limit reset, which refills every limit; the button isn’t in Claude Code, so it’s used from Claude in a browser or the desktop app.
• Upgrade to Max, or to a Premium seat.
• An API key from a Claude Console account, billed per token (next section).

Max and Team plans now come with monthly API credits ($100 on Max 5x, $200 on Max 20x, $20 or $100 a Team seat, pooled and capped at $500 a team), but they “don’t apply to” interactive Claude Code in the terminal, IDE, desktop or web, or to extra usage. They cover claude -p and the Agent SDK only when run with an API key from the linked Console organization; signed in with your plan, claude -p “still draw[s] from your plan’s usage limits,” per the credits article. They expire at the end of each billing cycle.

An API key instead of a plan

A Claude Console account runs Claude Code on prepaid API credits: “Claude Code consumes API tokens at standard API pricing,” with no five-hour or weekly limit. Anthropic’s costs page gives its own enterprise average as “around $13 per developer per active day and $150-250 per developer per month.” What a key doesn’t get is everything that needs a claude.ai sign-in: cloud sessions and the Claude app’s Code tab, Remote Control, the desktop app, routines, Slack, the Chrome extension and computer use.

The trap runs the other way. Claude Code takes an ANTHROPIC_API_KEY in the environment ahead of your plan’s login: interactively it asks you once, and in a non-interactive claude -p run “the key is always used when present.” Anthropic’s help article is blunt: “you’ll be charged via API pay-as-you-go rates” even if you’re logged in with a subscription. On a machine where a key is set for some other tool, unset ANTHROPIC_API_KEY returns Claude Code to your plan, and /status shows which one it is using.

On a server, or any machine without a browser

The plan’s login works over SSH: if the browser can’t reach Claude Code’s local callback, which the authentication docs say is “common in WSL2, SSH sessions, and containers,” you open the login link on any device and paste the code it shows at the Paste code here if prompted prompt. On Linux the login is kept in ~/.claude/.credentials.json, readable only by you, and Claude Code warns at startup when it is within three days of expiring; an unattended session that outlives it stops until you sign in again. For scripts and CI there is claude setup-token, a one-year token on the same plan that can make model requests but can’t start Remote Control.

Anthropic’s terms for Claude Code say that its rules do not prevent an end user from signing in to the unmodified Claude Code with their own Claude subscription, including where a platform hosts Claude Code. What else those terms and the consumer terms say about a plan on an always-on server is its own guide.

Where Anthropic’s own pages differ

As of October 9, 2026, Anthropic’s pages disagree with each other on a few points a buyer might act on. Each is stated as both pages put it:

• Fable on Max. The pricing page lists it at “50% of weekly limits,” while the help articles on Claude Code with Pro or Max and on usage credits (both updated October 8) say “Max plans also have a separate weekly limit for Fable.” The Max article itself mentions neither.
• Usage on Enterprise. Claude Code’s costs page says each Team and Enterprise member’s usage “draws from a per-seat allowance that resets on a rolling five-hour window and a weekly window,” while the Enterprise help articles say usage-based plans, the current single seat included, have “no per-seat usage limits.” The help articles call the older Standard and Premium Enterprise seats legacy.
• Cloud sessions on Enterprise. The cloud-session docs require “premium seats or Chat + Claude Code seats,” the older seat types; the help article on Claude Code for Team or Enterprise says Claude Code comes with the single Enterprise seat on new and self-serve plans, and the docs don’t mention that seat.
• Priority access on Pro. The pricing page marks “priority access at high traffic times” as Max only; the Pro help article lists “priority access to Claude during high-traffic periods” among Pro’s benefits.
• Team’s context window. The Team article says “200k context window”; the pricing page and the context-window article give up to 1 million tokens for the newest models, Claude Code included.
• Tax. The pricing page says “Prices shown don’t include applicable tax”; the Pro article says “some regions include applicable taxes in the displayed price while others add tax at checkout.”