Claude Pro vs Max (and Ultra): which plan do you need for Claude Code?

What are the Claude plans in October 2026?

Anthropic changed its individual plans during 2026, so older articles about “Max 5x” and “Max 20x” are out of date. As of October 2026 the Claude pricing page lists four individual plans. Anthropic’s Ultra plan help article confirms the rename: the Max 20x plan is now called Ultra, and existing Max 20x subscribers moved to Ultra automatically with no price change.

All figures exclude tax and, according to Anthropic, are for web subscriptions; mobile app store prices may differ. Anthropic also states that plans and prices can change at its discretion.

What do you actually get for the extra money on Max?

The Max plan help article lists the benefits. Compared with Pro, Max adds:

• Five times Pro’s usage in each five-hour window and in each week.
• Fable included, up to half of your weekly limit. Fable is Anthropic’s top model. On Pro it runs on pay-as-you-go usage credits instead.
• USD 100 a month in API credits to build on the Claude Platform. These credits cannot be spent in Claude Code or the Claude apps.
• Priority access at high-traffic times and early access to advanced features.

Everything else on the plan comparison is the same: Claude Code, Projects, research, Claude in Chrome, Microsoft 365 connections, memory and the Claude Design, Slides and Docs tools. Both plans reach Opus, Sonnet and Haiku. Max is not a different Claude Code. It is the same Claude Code with a larger allowance.

Source: the plan comparison on claude.com and the Fable plan article.

How do usage limits work for Claude Code?

Claude Code does not have its own allowance on these plans. Anthropic’s guide to Claude Code on paid plans says Pro and Max have a five-hour session limit and a weekly limit, and that Claude and Claude Code share them, so every chat and every coding session draws on the same pool. Max and Ultra also have a separate weekly limit for Fable.

• The five-hour window starts when you send your first message and resets in full five hours later.
• The weekly limit resets at a fixed time assigned to your account, not when you subscribe. You can see the next reset in Settings, under Usage.
• What affects usage: the length and complexity of conversations, the features you use, the model, and the effort level you choose. Longer conversations that trigger automatic context management consume more.
• Wrap-up allowance: if you hit the five-hour limit while Claude Code is partway through a response, it may keep working briefly to finish the step. Anthropic says this is rolling out gradually, and on Pro it applies up to once in each weekly usage period.
• Usage credits: on Pro, Max and Ultra you can switch on pay-as-you-go credits so work continues past a limit at standard API rates, with a monthly spending cap you set. They are billed separately from the subscription.

Anthropic does not publish limits as a number of messages or tokens. It describes each tier as a multiple of Pro, and the actual headroom depends on your habits, as set out in its guide to usage and length limits. The only way to know how far Pro goes for you is to measure it.

Is Max better value than Pro?

Take the published multiples at face value. Max gives 5x Pro’s usage for 5x the monthly price (USD 100 against USD 20). Ultra gives 10x Pro’s weekly usage for 10x the price (USD 200 against USD 20). If you pay Pro by the month, the cost per unit of usage is the same on all three tiers. If you pay Pro annually at USD 17 a month, Pro is the cheapest per unit by a small margin.

So the upgrade is not a discount. You are buying fewer interruptions, a bigger reserve for heavy days, Fable inside the plan, a monthly API credit for other projects and priority at busy times. That is worth it for someone who loses an hour whenever a five-hour window runs out. It is not worth it for someone who hits the limit once a month.

A caution on that arithmetic: Anthropic describes usage in multiples, not units, and says limits can be adjusted at its discretion, so the equal-value reading holds only while the published multiples hold. Re-check the pricing page before you commit.

Who should pay for which plan?

Team prices come from the same pricing page: a standard seat is USD 20 a month per seat billed annually and USD 25 billed monthly, and a premium seat with five times the usage is USD 100 billed annually and USD 125 monthly. Anthropic’s help articles describe Pro, Max and Ultra as plans for individual consumers and send organizations to the Team and Enterprise plans.

How do you decide with data instead of guessing?

Spend two weeks on Pro with a simple log. Settings, then Usage, shows the current session and your next weekly reset. Note when you hit a limit, what you were doing and how long you waited.

Date | Hours of Claude Code | Main task | Hit the 5-hour limit? (Y/N) | Time lost waiting (min) | Weekly usage shown at end of day (%)

At the end of week two, answer:
1. On how many days did I hit the five-hour limit? (Rule of thumb: 0 or 1 days, stay on Pro. 4 or more, price Max.)
2. Did I come close to the weekly limit before the reset? (Yes in both weeks: price Max.)
3. What did waiting cost me? (minutes lost x my hourly rate)
4. Would usage credits on Pro cover the spike days for less than the USD 80 a month gap between Pro and Max?
5. Do I need the top model or the monthly API credit, or only more time?

If the answers favor Max, upgrade for one month and repeat the log. Cancel back to Pro if the limit stops being the bottleneck.

Anthropic's help center also has usage-limit best practices. Smaller scoped tasks, a tight CLAUDE.md and new conversations for new topics all lower usage, and a clear build prompt avoids rework. Every idea on the board has a generated one, which is a cheap way to start a session with a scoped task.