Claude Max 5x costs $100 a month and Max 20x costs $200. Anthropic defines them as five and twenty times Pro's usage per five-hour session. Both tiers also have a weekly limit across all models and a separate Fable limit set at half of that week. Since October 7, 2026, Max 5x includes $100 and Max 20x includes $200 in monthly API credits.
The multiplier describes one session. Anthropic publishes no weekly figure for either tier, and the public evidence disagrees: its old 2025 estimates implied a 20x week under twice the 5x week, while one August 2026 test of three accounts found about 3.5 times. Start with 5x, and move to 20x when 5x sessions keep hitting the five-hour wall.
My LinkedIn post on the Claude Max lawsuit reached 61,573 impressions, 367 reactions and 173 comments. The case matters, and a buyer can still decide today without predicting it: measure accepted work before the weekly reset.
in
“A Claude Max customer sued Anthropic. The math is wild. Here are the 7 facts behind the lawsuit.”
Claude Max 5x vs 20x at a Glance (October 2026)
| Feature | Pro | Max 5x | Max 20x |
|---|---|---|---|
| Price per month | $20, or $17 billed yearly | $100 | $200 |
| Billing | Monthly or yearly | Monthly only | Monthly only |
| Usage per five-hour session | Pro baseline | 5x Pro | 20x Pro |
| Weekly limit, all models | Yes, size unpublished | Yes, size unpublished | Yes, size unpublished |
| Fable 5 and Fable 5.1 | Usage credits only | Up to 50% of the weekly limit | Up to 50% of the weekly limit |
| Monthly API credits | None | $100 | $200 |
| Claude Code | Included, shares plan limits | Included, shares plan limits | Included, shares plan limits |
| Usage credits after a limit | Yes, at API rates | Yes, at API rates | Yes, at API rates |
| Higher output limits | Not listed | Yes | Yes |
The multiplier lives in the session row. Anthropic publishes no weekly number as a multiple of Pro, so 20x describes the size of one session and leaves the size of your week open.
Every Claude plan, from Free to Enterprise, is compared in my Claude pricing guide.
Anthropic's Max plan help page defines both tiers. Check it and the pricing page before you buy, because the company says prices and limits can change. Both Max tiers also add higher output limits and priority access at high traffic times.
Claude Max Weekly Limit: Is 20x Really Four Times 5x?
Inside one five-hour session, Max 20x gives four times the room of Max 5x. For the week, Anthropic's help page says only that the limit applies across all models and resets at a fixed time. It gives no size for either tier. So the real 5x vs 20x question is the size of the week, and the public answers disagree.
| Source | Date | What it measures | 20x week vs 5x week | Status |
|---|---|---|---|---|
| Anthropic Max plan help page | Read October 12, 2026 | Usage per five-hour session | 4x per session; weekly ratio unpublished | Official |
| Anthropic weekly estimates reported by TechCrunch | July 28, 2025 | Weekly hours of Sonnet 4 and Opus 4 in Claude Code | About 1.7x for Sonnet 4 and 1.3x for Opus 4, at range midpoints | Old official estimate, gone from the help center |
| Kahn v. Anthropic complaint | June 14, 2026 | Plaintiff's estimate of usage against Pro | Alleges 6x to 8x Pro for 20x and about 3.5x Pro for 5x | Allegation, unproven |
| Classmethod test of three accounts | August 2026 | Weekly meter readings matched to local Claude Code logs | About 3.5x, range 3.1x to 3.9x | Third-party test by an Anthropic reseller, run during a temporary Claude Code boost |
I read the table this way. Nothing public confirms a 4x week. The complaint's multiples match the arithmetic of Anthropic's 2025 hour estimates, so those two rows rest on the same old numbers. The only recent measurement I found points to about 3.5x. Until Anthropic publishes a number, I would plan a 20x week at about twice a 5x week and treat anything above that as a bonus.
The five-hour clock controls intensity: can a demanding session keep moving right now? The weekly clock controls endurance: how much included work the account carries before its assigned reset. You can have room in the next session and still be stopped by the week, so a larger session ceiling is a different thing from a guaranteed week of heavy work.
What Claude Max Usage Limits Actually Mean
One pool covers Claude web, desktop, mobile and Claude Code, so a long coding job, a research project and ordinary chats consume one allowance. Anthropic also says it may limit usage in other ways, such as monthly caps or model and feature limits. A team that budgets each interface separately will see its real capacity too late.
There is no fixed message count. Anthropic says consumption depends on conversation length, files, the model, the features you use and the effort level you pick. Ten short requests can cost less capacity than one agent run over a large repository with tool calls. The practical unit is completed work under the reset rules.
The weekly limit resets at a fixed day and time assigned to your account, whenever you start using Claude. On Max, Settings, Usage shows three bars: the current session, the weekly limit for all models and the weekly Fable limit. In Claude Code, the /usage command shows your limits.
What you may use Claude for is covered in my Anthropic usage policy guide.
Claude Max Fable Usage: The 50% Rule
Fable 5 and Fable 5.1 are included in both Max tiers, with one cap. Max users can spend up to 50 percent of the weekly limit on Fable at no extra cost. That ceiling sits inside the same pool, and the total can never exceed the weekly limit. Anthropic also says Fable uses your limits faster than other Claude models.
Past the 50 percent mark you either keep using Fable with usage credits or switch to another model. On Pro, Fable sits outside the plan's limits altogether and runs on usage credits from the first message. If Fable is the reason you work in Claude, that difference alone can justify Max.
Model choice is now a cheap lever. Anthropic says Opus 5.5, launched on September 22, performs at the level of Fable 5.1 on most work. In the Classmethod test, Fable 5 moved the weekly meter about twice as fast as Opus for the same API value. On Max I make Opus 5.5 the default and keep Fable for the jobs where it clearly wins.
Anthropic's Fable models on your plan page explains the 50 percent rule, what happens at the cap and how Pro differs. The earlier promotion that put Fable 5 inside Pro limits ended on July 19, 2026.
Claude Code Limits on Max: What Changed in 2026
Anthropic's pricing page still says Claude on the web, desktop, mobile and Claude Code draw from one pool. Even so, it has adjusted Claude Code limits on their own three times this year, for Max 5x and Max 20x alike. Check the dates before you trust any weekly-limit test, including the one in the table above.
| Date | Change | Effect on Max |
|---|---|---|
| May 6, 2026 | Claude Code five-hour limits doubled; peak-hours reduction on Claude Code removed for Pro and Max | More room per session in Claude Code on both tiers |
| May 13 to September 13, 2026 | Promotion: Claude Code weekly limits 50% higher | Bigger Claude Code week; chat and Cowork limits unchanged |
| September 14, 2026 | Claude Code weekly limits set 25% above their level before the promotion | Anthropic called it a 17% cut against the promotion |
| October 7, 2026 | Monthly API credits added | $100 on Max 5x, $200 on Max 20x |
The practical lesson: a weekly-limit test run in August 2026 measured a bigger Claude Code week than the one you buy today. Since September 14, the Claude Code week sits 25 percent above its level before the promotion and, by Anthropic's own math, about 17 percent below the promotion.
Usage Credits, API Credits and Limit Resets
Max now carries three kinds of extra capacity, and they are easy to confuse. Usage credits pay for work past your limits. Monthly API credits pay for apps and agents you build on the Claude Platform. Limit resets are occasional free refills. Of the three, only usage credits and limit resets help when Claude itself stops at a limit.
| Option | What it covers | Cost | Catch |
|---|---|---|---|
| Usage credits | Work in Claude, Claude Code and Cowork after the session, weekly or Fable limit runs out | Standard API rates, prepaid, with an optional monthly spend limit | Billed on top of the plan; turn on in Settings, Usage |
| Monthly API credits | Claude API, Managed Agents, the Agent SDK and the Console playground | Included: $100 on Max 5x, $200 on Max 20x | Exclude interactive Claude Code and extra usage; expire each billing cycle; claimable after seven days on the plan |
| Limit resets | Refill the five-hour or the weekly limit to full | Free when Anthropic offers one | Occasional; used on web or desktop; each offer expires |
The API credits change the 5x vs 20x math only for builders. Max 20x costs $100 more than Max 5x and includes $100 more in monthly API credits. If your company would spend that much on the Claude API anyway, the real gap between the tiers shrinks. If you never call the API, the credits cannot pay for Claude Code or extra usage, and the gap stays $100.
What the Claude Max Lawsuit Alleges
Karl Kahn filed Kahn v. Anthropic PBC in the Northern District of California on June 14, 2026. On September 4 the court consolidated it with Smith v. Anthropic, and the plaintiffs filed a consolidated complaint on September 8. Anthropic moved to dismiss on September 22, and the court set the hearing for November 6, 2026.
The original complaint alleges that Max 20x delivers six to eight times Pro's usage and Max 5x about three and a half times. It also describes one five-hour session that used about 15 percent of the plaintiff's weekly allowance. These are allegations from one side of an active case. No court has ruled on them.
The public case docket lists each filing, including the consolidation and the motion to dismiss. Neither the docket nor the complaint establishes that the claims are true.
A Buyer Test for Claude Max Usage Limits
Run one representative week on your current tier before you change plans. The goal is a cost per accepted job, measured on your own work instead of a feature list.
- Pick ten recurring jobs that justify the plan, such as code changes, account research, analysis, document review or agent runs. Use real inputs, including long contexts and tool calls.
- Write down what an accepted output looks like before you start, so fluent drafts do not inflate the result.
- For each job, record the surface, model, effort level, session start, completion state and human review time.
- Read Settings, Usage before and after each heavy job and note the session, weekly and Fable percentages.
- At the end of the week, count accepted outputs, interruptions, jobs postponed by a limit and usage credits spent.
- Divide subscription and credit spend by accepted jobs. If the five-hour wall caused most interruptions, price Max 20x. If the weekly wall did, price usage credits first.
Repeat the same sample after a model, limit or plan change, with a frozen baseline and every output stored. Both habits come from my regression testing routine.
Max 20x vs 5x: Where the Math Breaks
The first failure is treating 20x as four times the weekly work of 5x. Anthropic's definition supports that ratio inside a five-hour session only. The weekly evidence above runs from under 2x to about 3.5x, so I would keep 4x out of any capacity plan.
The second failure is using a consumer subscription for predictable production throughput. A subscription is convenient for interactive work. A metered API path offers clearer unit pricing and controls for automated systems, though it adds engineering and variable spend. If the business needs guaranteed service characteristics, compare Team or Enterprise terms instead of inferring them from Max.
If you are weighing that metered path, my breakdown of Claude Opus API pricing shows what an agent session costs on Opus 5.5 with and without caching.
For a side-by-side of those team plans, read my Claude vs ChatGPT for business comparison.
OpenAI went the other way: its $200 plan reopened on September 29 with half the old usage. My page on ChatGPT Pro 20x, now Pro 200 at 10x works out the price per unit of usage across its tiers.
The Decision Rule
Buy Max 5x when Pro keeps interrupting accepted work or when you need Fable inside the plan. Move to Max 20x when 5x sessions still hit the five-hour wall on work that matters, or when your company would use the extra $100 in API credits anyway. If the weekly limit or the Fable cap stops you, price a week of usage credits first.
I would review the decision monthly with three numbers: accepted jobs, interruption hours and total spend including credits. That scorecard turns Claude Max usage limits into an operating choice. The lawsuit asks whether the marketing created a misleading expectation. Your own ledger answers the question that matters inside the company: what did the plan reliably help finish?
Use the AI model evaluation framework to build representative cases and acceptance criteria before comparing plans or providers.
Every Wednesday, AI Frontier gives B2B operators one verified AI signal and one practical play. If you track your Claude work this week, send me the one job that consumes more allowance than you expected.
Frequently asked questions
What is the difference between Claude Max 5x and 20x?
Max 5x costs $100 a month and Max 20x costs $200. Anthropic defines them as five and twenty times Pro's usage per five-hour session. Both have a weekly limit across all models, a Fable limit at half of that week and monthly API credits of $100 or $200. Anthropic publishes no weekly ratio between the two.
Is Claude Max 20x four times Max 5x per week?
Nothing public confirms it. Anthropic states the 4x ratio per five-hour session only. Its old 2025 estimates implied under 2x per week, the figures in the Kahn complaint imply roughly 1.7x to 2.3x, and one third-party August 2026 test of three accounts measured about 3.5x. Check your own Settings, Usage readings before you plan on more.
Does Claude Max include API credits?
Yes, since October 7, 2026. Max 5x includes $100 and Max 20x $200 a month for the Claude API, Managed Agents, the Agent SDK and the Console playground. You claim them by linking a Claude Console organization after seven days on the plan. They expire at the end of each billing cycle and leave your Claude usage limits unchanged.
Do Max API credits work with Claude Code?
They exclude interactive Claude Code in the terminal, IDE, desktop or web, and they exclude extra usage. Anthropic says they do cover claude -p and the Agent SDK when you run them with an API key from your linked Console organization. Signed in with your plan instead, claude -p draws from your plan limits.
Is Fable included in Claude Max?
Yes. On Max 5x and Max 20x you can spend up to 50 percent of your weekly limit on Fable 5 and Fable 5.1 at no extra cost, and Fable uses the limits faster than other models. After the cap, you continue with usage credits or switch models. On Pro, Fable runs on usage credits from the start.
Why can't I use Fable in Claude Code on Max?
Check your version first. Anthropic lists Claude Code 2.1.170 or later for Fable 5 and 2.1.255 or later for Fable 5.1. If you have already used 50 percent of your weekly limit on Fable, Fable runs on usage credits until your weekly reset.
When do Claude Max usage limits reset?
On two clocks. The session limit resets every five hours. The weekly limit for all models resets at a fixed day and time assigned to your account, whenever you start using Claude, and the Fable limit also resets once a week. You can see the next reset in Settings, Usage.
How many messages do you get on Claude Max?
Anthropic publishes no message count. Usage depends on conversation length, files, model, features and effort level. Its older Max usage article, which many guides still quote for message estimates, now returns an error page.
Is Claude Max unlimited?
No. Max raises the session allowance, and weekly limits still apply. Claude web, desktop, mobile and Claude Code draw from the same pool, and Anthropic may add other caps such as monthly or feature limits.
What happens when I hit my Claude Max limit?
You can wait for the reset time Claude shows, use a limit reset if one is available, move to a higher plan or turn on usage credits. Credits are billed at standard API rates, and you can set a monthly spend limit in Settings, Usage. Monthly API credits cannot pay for this extra usage.
Can I pay for Claude Max yearly?
Max is billed monthly only. Pro offers yearly billing at $17 a month, paid as $200 up front. If you move from Max 5x to Max 20x, Anthropic charges the difference pro rata for the rest of your billing cycle.
Can I get Claude Max 20x for free?
Anthropic's Claude for Open Source program offers six months of free Max 20x to qualifying maintainers and contributors. Outside programs and occasional promotions, Anthropic says it offers no standing discounts on its paid plans.
Should I buy Max 20x or use usage credits?
Measure first. If the five-hour window interrupts valuable work most days, the larger session allowance of Max 20x protects it for a flat fee. If limits bite only a few times a month, usage credits with a spend limit can cost less. Log accepted jobs for a week before you choose.
