Claude's new plans: Max, Ultra and what they mean for teams

At the end of October Anthropic renamed its two top personal plans and changed how limits work on the biggest one. If your engineers use Claude Code all day, this decides whether a long agent run stops in the middle of the afternoon, and whether you are paying for capacity nobody uses. The change itself is small. The questions it raises in companies are bigger: which seats for whom, what the new API credits actually cover, and why personal plans are the wrong answer for company work even when they look cheaper. Here are the facts from Anthropic's pricing page and help center, and how I would handle them in a team.

Dominik Heinrich

November 7, 2026Updated November 8, 20269 min read

1. What changed on 30 October

Max 5x is now simply called Max, and Max 20x is now called Ultra. Anthropic's help center states it plainly: the Max 20x plan is now called the Ultra plan, existing subscribers move over automatically and the price doesn't change. Max stays at 100 dollars and Ultra at 200 dollars a month; according to The Next Web's coverage, the European prices are 90 and 180 euros. Most existing subscribers got the change on Friday, 30 October. Both plans are billed monthly only, while Pro is available monthly at 20 dollars or annually at 17 dollars a month. Nothing changed on the Team and Enterprise side. The rename matters more than it looks, because the old names described the five-hour window, and that window is exactly what changed on the bigger plan.

2. No five-hour window on Ultra, but still a weekly cap

Most Claude accounts have a limit that resets on a rolling five-hour window, and paid plans add a weekly limit on top. Activity across the web, desktop, mobile and Claude Code all draws from the same pool. Verified Ultra accounts no longer have the five-hour limit, only the weekly one: 10 times Pro's usage per week. Max keeps both, at 5 times Pro every five hours and every week. So the old '20x' described the five-hour window, while per week Ultra gives 10 times Pro. Verification is automatic, based on account and payment history; verifying your phone number at checkout is a separate step and doesn't remove the limit. You can check under Settings, Usage: if there is no 'Current session' bar, your account is verified. Ultra also has a separate weekly limit for Fable, and Anthropic reserves the right to apply further caps to manage capacity. For agent work this is the real change: a long /goal no longer stalls after a few intense hours, as long as the week has room.

Tip from practice

If you hit the weekly limit anyway, a limit reset (when Anthropic offers one under Settings, Usage) restores the weekly limits to full. Save it for a migration week, not a Tuesday.

3. The plans side by side

According to Anthropic's pricing page as of November 2026: Free costs nothing and has limited usage. Pro costs 20 dollars a month, or 17 dollars a month billed annually, includes Claude Code and pays for Fable with usage credits. Max costs 100 dollars a month, gives 5 times Pro every five hours and every week, includes Fable for up to half the weekly limit, 100 dollars a month in API credits and priority access at high-traffic times. Ultra costs 200 dollars a month, has no five-hour limit for verified accounts, 10 times Pro per week, Fable for up to half the weekly limit and 200 dollars a month in API credits. Team, for 2 to 150 people, costs 20 dollars per standard seat and 100 dollars per premium seat per month on annual billing (25 and 125 dollars monthly); a premium seat gets 5 times a standard seat's usage, and the team gets SSO, central billing, admin controls for connectors and up to 500 dollars a month in pooled API credits. Enterprise costs 20 dollars per seat per month, billed annually with a 20-seat minimum, plus usage at API rates, and adds spend limits, role-based access, SCIM, audit logs, a compliance API and custom data retention. Compare them in the table below.

Diagram of the Claude plans after 30 October 2026: personal plans Free, Pro, Max (5x Pro) and Ultra (10x Pro) with their weekly usage compared with Pro, and company plans Team and Enterprise, where only the company gets SSO, central billing, connector controls, managed Claude Code settings and usage analytics. — Individuals get capacity. The company keeps control. Source: Anthropic pricing page, November 2026.

Interactive

Which Claude plan fits how you work?

Pick the description closest to you. Plans and prices as of the change on 30 October 2026 (USD), from Anthropic's pricing page and help center.

Max

Price: $100/month (monthly only)
Usage: 5× Pro per 5 hours and per week
Also included: $100 API credits a month (API, Agent SDK, not Claude Code); Fable up to half the weekly limit

All calculators and checklists

4. What the API credits are (and aren't) for

The new monthly credits are easy to misunderstand. Anthropic's platform docs list what they cover: the Claude API, Claude Managed Agents, the Claude Agent SDK and the playground. They don't cover Claude Code or extra usage in the Claude apps, so they don't stretch your agent sessions. You claim them by linking a Claude Console organization to your plan, they arrive each billing cycle and expire at its end with no rollover, and new subscribers can claim them after seven days. On Team, credits of 20 dollars per standard seat and 100 dollars per premium seat are pooled into one balance capped at 500 dollars a month; Free, Pro and Enterprise aren't eligible. In practice they are the budget for everything around the agents: an evaluation suite that scores prompts, a small internal tool built on the Agent SDK, a nightly script that summarises incidents. Used that way, 100 or 200 dollars a month covers a surprising amount of glue work that would otherwise need a separate API budget and approval.

Tip from practice

Credits expire every cycle, so give them a standing job (evals, reports) instead of saving them for 'a project later'. Unused credit is the only kind of waste here that nobody notices.

5. Plan heavy runs around the week, not the month

Usage refills on rolling windows, not on your billing date, and how fast you reach a limit depends on the model, the length of the conversation and the size of the files. That changes how I plan big jobs. A large /goal, for example a framework migration across a monorepo, should start early in the week with a turn clause that fits the remaining allowance, not on Friday afternoon. In Claude Code, /usage shows session cost, plan usage limits and activity stats, including a breakdown of what counts against your plan, so check it before you start something that runs for hours. And when a goal does hit a usage limit, the Claude Code docs describe that it pauses with a notice naming the cause instead of failing silently, so you can resume when the window frees up. Fable is the expensive part: on Max and Ultra it can use up to half the weekly limit, so I keep it for the hardest tasks and run routine work on Opus or Sonnet.

text

/usage
# Check the weekly allowance first, then size the goal to it:
/goal the monorepo builds with the new framework version
(npm run build exits 0 and npm test exits 0 in every package),
no package.json outside packages/ changes, no test is skipped.
Or stop after 40 turns and list the packages that still fail.

6. Why personal plans are the wrong answer for company work

It's tempting to expense a few Ultra subscriptions for the heaviest users and be done. I advise against it. Anthropic's own comparison table lists SSO, central billing, admin controls for connectors, role-based access, SCIM, audit logs, usage analytics and the compliance API as unavailable on Free, Pro, Max and Ultra. In practice that means no offboarding when someone leaves, no central view of what the agents did, and company code sitting in personal accounts. On Team and Enterprise, admins can also push Claude Code managed settings that users can't override, through a file, MDM or the admin console: deny rules for secrets, no bypass mode, an approved list of MCP servers and the models people may select. For heavy users, Team has premium seats with 5 times the usage of a standard seat; Enterprise has usage at API rates with spend limits per user and per organisation. Both give you the capacity of the personal plans without losing control of the code.

json

{
  "permissions": {
    "deny": ["Read(./.env)", "Read(./secrets/**)"],
    "disableBypassPermissionsMode": "disable"
  },
  "availableModels": ["opus", "sonnet", "haiku"]
}

7. How I would assign seats

Start everyone who writes code on a Team standard seat, and give premium seats to the few people who run long agent sessions every day: platform engineers, the people doing migrations, whoever owns the nightly routines. Use the pooled API credits for evaluations and internal tools instead of separate API keys on personal cards. Move to Enterprise when you need SCIM, audit logs, spend limits or custom retention, which in regulated companies is usually from the start. Then look at real usage every month. In most teams I've seen, a handful of people use most of the capacity and many seats are barely touched, so the right mix changes after the first month. The plans will keep changing names and numbers, as they just did. The principle doesn't: individuals get capacity, the company keeps control.

Takeaway
Ultra makes long autonomous runs practical for heavy users; for a team, the bigger win is central control: capacity for individuals, managed settings and audit for the company.

Real projects behind this

From AI prototype to a production SaaS

Try it yourself

Claude plan finder

Want this in your company? Let's talk mail@dominikheinrich.ch

Sources and official guides