Claude Pro limits are, according to Anthropic, at least five times the usage per five hour session compared with the Free plan, and Max raises that to five or twenty times Pro usage, depending on the tier (source: support.claude.com, as of June 10, 2026, and support.claude.com, as of August 7, 2026). What those numbers don't tell you: how the session limit interacts with an additional weekly limit, and a genuine contradiction in Anthropic's own sources about how Fable models get billed on Pro. I'll clear up both.
I pay for several Claude subscriptions at once, because my AI workforce is spread across Pro and Max accounts. I collect all my Claude guides on the Claude overview page. The prices themselves, Free through Team, are already laid out completely in Claude Pricing and Plans Explained; this article is only about the question that comes after: Pro or Max, and how you notice the difference day to day.
Claude Pro limits: the table with Max 5x and 20x
The three tiers differ mainly in the multiple of usage, not in the core features. All three include, according to Anthropic, Claude Code, Cowork, Projects, and Opus access, which makes Pro the entry point for everything described on this site.
| Plan |
Price per month |
Usage per session |
Fable models |
| Pro |
20 dollars (17 annually) |
at least 5x Free |
via usage credits from the first message (per help article) |
| Max 5x |
100 dollars |
5x Pro |
up to 50% of weekly limit included, then usage credits |
| Max 20x |
200 dollars |
20x Pro |
up to 50% of weekly limit included, then usage credits |
Prices per claude.com/pricing, as of September 2026. Max also adds priority access to new models (source: support.claude.com, as of August 7, 2026). The full table with Free and Team is in Claude Pricing and Plans Explained.
How the session limit and the weekly limit interact
A misunderstanding I run into often: people budget only for the five hour window and then wonder why Claude slows down earlier anyway. Anthropic explicitly names two separate limits for Pro: a session limit that resets every five hours, and an additional weekly limit on top of it (source: support.claude.com, as of June 10, 2026). Both apply at the same time. You can still have quota left within a five hour window and still hit the weekly limit if you've worked heavily on the days before.
Anthropic doesn't publish concrete figures for either limit; only the usage display in your own account settings shows the current state, according to Anthropic. So if you're trying to plan how far a plan will get you, don't just think in terms of one long day, think in terms of the whole week.
The Fable contradiction: what I can actually tell you
While researching this article, I ran into a real contradiction in Anthropic's own sources, and I'd rather tell you that honestly than paper over it. The help article on Fable models describes it like this: on Max and Team Premium, Fable models are included up to 50% of the weekly limit, after which usage credits kick in. On Pro and Team Standard, by contrast, Fable models run on usage credits from the very first message, with no included quota at all (source: support.claude.com).
The pricing page at claude.com/pricing, however, also shows a "Fable (50% weekly limits)" note for Pro elsewhere, which contradicts the help article. I've read both sources carefully and found nothing that clearly confirms which one is current. In my own account, in September 2026, I see Fable inside the weekly limit. That is a single observation on one account at one point in time, not evidence of how Anthropic bills this in general, and it does not resolve the contradiction in the sources. So instead of inventing an answer, here's my honest advice: check this directly in your own account under usage before you make a purchase decision, and if you're going to trust one source, use the Anthropic help article on Fable models as your reference, it's the more detailed description of the two cases.
How you notice Pro isn't enough anymore
A single limit warning isn't a reason to upgrade on its own. What tipped it for me was a recurring pattern: several times in the same week, the usage display showed me hitting the session limit before my workday was actually done, not just once on an unusually busy day. The second signal was structural: tasks that need several Claude Code sessions running in parallel, say because multiple projects are moving at once, hit Pro's ceiling faster than a single chat ever would.
If both patterns apply to you, switching to Max is worth it. If only one does, I'd watch for another week before upgrading.
Max or Team Premium instead?
If you work alone, there's not much way around Max. But once several people in an organization are working with Claude, it's worth comparing against Team Premium: a Standard seat gives, according to Anthropic, 1.25x Pro usage per session, a Premium seat 6.25x, plus central administration and a privacy commitment that content isn't used for training by default (source: support.claude.com). For one person doing a lot of work, Max stays the simpler choice. Once a team is involved, Team Premium is often the more economical option per head.
My workday: an AI workforce in practice
I'm deliberately not naming a token count here, because I don't have a documented one and any estimate would depend on my own way of working in a way that doesn't transfer to you. Instead, here's what a workday actually looks like for me.
On a normal day, not all eight employees are active at once, but several nearly always overlap. Conny works on content drafts while, in parallel, Sebastian runs an audit or Ralf sits with a website analysis, each session its own Claude Code run with its own context. Some sessions are short and targeted, a single piece of text or a single evaluation. Others stretch over hours, with many file operations, intermediate checks, and correction loops, for example when several blog posts get researched and written back to back.
That overlap is exactly why I work on Max today, not any single calculated figure. During a stretch with a lot of knowledge graph work, two Max plans ran side by side for a while. Once several roles are working long and in parallel on the same day, the Pro quota fell short for me regularly, not just on exceptional days. If you're just getting started with one or two AI employees, this pattern probably isn't relevant to you yet, and Pro is usually enough.
Frequently asked questions
How much more do I get with Max versus Pro?
According to Anthropic, at least five times the usage with Max 5x and twenty times with Max 20x, each measured against Pro. Anthropic publishes no fixed usage numbers beyond these multiples.
Can I cancel Max monthly?
I won't state a cancellation policy without a source for it. The reliable answer is in your own account under subscription settings.
Is Opus included in Pro?
Yes, according to Anthropic, Opus access is part of the Pro plan, along with Claude Code, Cowork, and Projects. The difference from Max is the usage multiple, not the feature set.
Is Max worth it specifically for Claude Code?
If you use Claude Code for long, file-heavy sessions or run several projects in parallel, you'll hit Pro's ceiling faster than in plain chat. For how Claude Code's costs actually break down, see Claude Code Pricing: Plan or API?.
What happens when I hit the limit?
You wait until the rolling five hour window or the weekly limit resets, then your quota is available again. If you need to keep working past that, paid plans let you use usage credits, which you manage yourself (source: support.claude.com).
Where to go from here
Start by watching on Pro whether the session limit or the weekly limit actually slows you down before switching to Max. If you're still unsure whether you even need a paid plan, the honest answer to that is in Claude Free Plan: What You Get.
The eight AI employee packages, built on the same principles as my own workforce, are available ready-made in my community.
