🚀 Claude Ultra Plan 2026: Limits, Pricing & Changes
Discover Claude Ultra plan changes for 2026: five-hour cap removed, verified limits, and Claude Max pricing. See if Anthropic Claude fits your needs.
Claude Ultra Removes the Five-Hour Limit. What Subscribers Need to Know
A subscription can feel "unlimited"... right up until a usage cap kicks in mid-afternoon and you're suddenly waiting for access to reset. That's the practical issue behind Anthropic's latest update to Claude's paid plans: it renamed the two Max tiers and removed the five-hour limit for verified accounts on the more expensive plan.
The $100-a-month Max 5x plan is now called Max. The $200-a-month Max 20x plan is now called Ultra. Prices haven't changed, but Ultra subscribers who are verified no longer hit a usage limit that resets every five hours. The weekly limit still applies, and some newer accounts may continue to see a temporary five-hour limit until Anthropic finishes verification.
This matters, because Ultra is not "unlimited Claude access." The new name also doesn't tell you exactly what you'll get day to day. Below is what changed, how the remaining limits work, what businesses should think about before relying on Ultra, and what this update suggests about subscription-based AI tools.
What changed in the Claude Max and Ultra plans?
Anthropic simplified the names of its two Max subscriptions. The former Max 5x plan is now Max, and the former Max 20x plan is now Ultra. Monthly prices stay the same: Max costs $100 in the US, while Ultra costs $200. In Europe, the listed prices are €90 and €180 per month, respectively.
The actual change is for Ultra. Verified Ultra accounts no longer have the five-hour usage limit that used to determine when their usage would resume. Most existing subscribers got the update when Anthropic announced it. Newer accounts may still have the temporary limit until Anthropic verifies them, and the company says it will email those users once the restriction is lifted.
The lower-priced Max plan still includes a five-hour limit. Anthropic's pricing page describes it as five times Pro usage, tracked over five-hour and weekly periods. Ultra is listed as ten times Pro usage per week, with no five-hour limit for verified accounts. Both plans also include access to the company's top model, identified in the pricing information as Fable, for up to half of the weekly limit.
One thing to watch: a plan name isn't the same thing as a guaranteed usage policy. The old "20x" label implied a simple comparison with Pro. The current Ultra listing still ties it to Pro - just as "ten times Pro usage per week." Anthropic's announcement doesn't spell out why that numeric framing changed, so you shouldn't assume the new name preserves the old comparison. What you can actually rely on are the current usage windows and limits, not the multiplier from the former tier name.
For existing subscribers, Anthropic says there's nothing to do. The price is unchanged, and the account update is handled by the company. If you have a newer account and still see a five-hour restriction, the best move is to wait for the verification email - don't treat the plan as immediately unrestricted.
In other words: it's a rebrand plus a limit tweak, not a price drop, not a move to "unlimited," and not a change across every tier. Whether Ultra feels better depends on which plan you're on, whether your account is verified, and how much your work gets disrupted by short usage windows.
How do Claude Ultra usage limits work now?
The simplest way to think about Ultra is to split its limits into two time windows: a short five-hour window and a weekly allowance. Anthropic removed the five-hour limit for verified Ultra accounts. It didn't remove the weekly limit. So you can avoid the five-hour reset - while still having a finite amount of usage across the week.
And it's worth saying plainly: "no five-hour limit" doesn't mean "no limits." It only removes that specific five-hour cap for verified accounts. Anthropic still frames Ultra around a weekly usage allowance, and that's the constraint that remains.
Verification is part of the plan experience
If you're buying Ultra for time-sensitive work, verification isn't a minor detail. Anthropic says most existing subscribers already got the change, while newer accounts may keep the five-hour limit until verification is complete. The company says it uses this step to reduce fraud and abuse, and it will email users when the restriction is lifted.
For an individual, a temporary limit can be annoying. For a team, it's something you should test before you build a workflow around "always-on" access. Don't assume a brand-new Ultra account behaves the same as a verified long-standing one from day one.
Weekly capacity is the planning constraint
For a team evaluating Ultra, the weekly limit is the real boundary. The relative usage number helps for comparison, but it won't tell you how many tasks or hours your specific workload will support. Usage can swing based on things like how much context you're feeding the model, how long conversations run, which models you use, and how often work repeats.
A practical check is to run a representative week, not a quick demo. Pay attention to who uses the account, what types of work burn the most capacity, when the team hits restrictions, and whether usage concentrates in a few heavy sessions. A plan that feels generous during a light week can become restrictive when you're dealing with a launch, a client deadline, or a major code review.
Ultra also includes monthly API credits: $100 for Max and $200 for Ultra, per Anthropic's pricing info. Those credits are separate from the subscription's usage allowance. If you're budgeting, treat them as a different component and verify the current terms before you plan around them.
The core takeaway is straightforward: view the five-hour change as more scheduling flexibility - not more total capacity. If your work regularly runs into the weekly allowance, removing the five-hour reset won't magically solve the underlying limit.
Is Claude Ultra worth $200 a month for a business?
For most people, the decision isn't about the headline price - it's about the cost of interruptions. Ultra costs twice as much as Max, so the real question is whether dropping the five-hour restriction for verified accounts saves enough time to justify the extra $100 per month.
For a solo professional, this can be pretty simple. If you're using Claude intermittently for drafting, research, or coding help, Max's five-hour window may be fine. If you rely on Claude during long, focused sessions, avoiding the five-hour reset can be genuinely useful - though the weekly limit still applies.
For a business, evaluate Ultra against your actual workflow, not an imagined "productivity boost." Pick something you do regularly - reviewing support conversations, drafting an initial write-up, helping a developer work through a codebase. Track how often the assistant gets used, how much human review remains, and what happens when Claude access is unavailable. Then compare that operational cost to the subscription fee.
A practical decision test
Ask these questions before upgrading:
- Are five-hour interruptions a recurring problem? If not, Ultra's main advantage may not matter much to your team.
- Does your workload fit within the weekly limit? If you already hit weekly restrictions, removing the shorter window won't be enough.
- Is the account verified? Newer accounts may still face the temporary limit, so don't promise "continuous access" before verification finishes.
- Can the work tolerate a fallback? If a task must continue when Claude is down, set up a manual process or alternative tool instead of treating the subscription like guaranteed infrastructure.
- Have you measured the benefit? Compare time saved and interruptions avoided against the extra $100 per month over Max.
That last part is often where the math becomes obvious. If waiting for a reset costs real billable time - or if production work keeps getting paused - Ultra can make sense. If the benefit is just occasional convenience, Max may be the better buy.
Also, separate a personal subscription from a managed business setup. A consumer plan can work for individual use, but a company should consider whether it needs shared administration, predictable availability, data controls, or integration with internal systems. The details here cover usage and pricing; they shouldn't be treated as the full answer to procurement questions.
In real projects, the safest approach is a limited trial with a real workload, keep a fallback for important work, and reassess after a billing period that looks like your normal usage. Upgrade when it removes a problem you've actually hit - not just because a higher tier sounds better.
What does the plan change signal about AI subscriptions?
Anthropic's update reflects a tension baked into paid AI products: customers want dependable access, while providers have to manage limited computing capacity and limit abuse. A five-hour reset can make capacity management easier, but it can also interrupt work at the exact moment a user expects a paid tool to behave like a reliable service.
Removing the short-window limit for verified Ultra accounts increases flexibility. Instead of tying access to a recurring five-hour boundary, users get more control over when they spend their weekly allowance. Still, because the weekly limit remains, the change reshapes the constraint rather than eliminating capacity management entirely.
That distinction matters for buyers across the AI market. Subscription plans are often described with multipliers, model names, or broad claims about access. Those labels are easy to compare on paper and hard to translate into a reliable operational budget. The details that actually affect your day-to-day work are usually more specific: what time period counts, which models apply to the allowance, whether verification is required, and what happens once you hit the cap.
Anthropic has also adjusted Claude's limits before. Reports say that in May the company doubled five-hour limits for Claude Code, citing new computing capacity from SpaceX. They also say that in April Anthropic blocked OpenClaw and other third-party agent tools from using Claude subscriptions. These examples show that access terms can shift when providers respond to infrastructure, cost, and abuse concerns. They don't guarantee future changes will always favor subscribers.
There's also a trust angle when plan names and usage descriptions stop lining up. The old "20x" label is now replaced by a listing that describes Ultra as ten times Pro usage per week. A lawsuit filed in June alleged that Anthropic oversold usage on both Max plans; that's an allegation, not something established by the information here. Still, when labels change and expectations get disputed, it's reasonable for customers to review the live plan terms instead of relying on old comparisons or marketing shorthand.
For teams, the broader lesson is operational: don't build your workflow so tightly around a subscription that it becomes a single point of failure. Keep key processes reviewable by people, retain a fallback for time-sensitive tasks, and monitor real usage. As AI assistants move from experiments into daily work, predictable limits and clear terms start mattering as much as model capability.
Key takeaways
- Max is the renamed $100 plan; Ultra is the renamed $200 plan.
- Verified Ultra accounts no longer have the five-hour usage limit.
- Ultra still has a weekly usage limit, and newer accounts may need verification.
- The former 20x label should not be treated as a current usage guarantee.
- Upgrade based on measured interruptions and workload, not the plan name alone.
Conclusion
Claude Ultra's update is helpful if your work keeps getting cut off by a five-hour cap, but it doesn't turn a paid plan into unlimited access. Verification still matters for newer accounts, and the weekly allowance remains the bigger capacity boundary.
If you're deciding whether to pay for Ultra, start with your actual usage: identify where the five-hour reset costs you time, check whether weekly limits are already a problem, and compare the real benefit to the extra $100 per month. Treat the plan as a tool with defined limits - not guaranteed infrastructure - and pick the tier that matches the work you need it to do.