Site icon Decode the future

Claude Ultra Plan: Max 20x Renamed, New Limits Explained

Cut-paper illustration of stacked subscription cards, a usage gauge and an abstract code window.

Last updated: November 7, 2026 · Plans, prices and limits checked against Anthropic’s pricing page, Help Center and Claude Code docs on this date

Claude Ultra is the new name for Anthropic’s $200-a-month Max 20x plan, renamed on October 30, 2026 with no price change. It gives 10 times Pro’s weekly usage, and verified Ultra accounts no longer have the rolling 5-hour limit. The weekly limit still applies, and Claude chat and Claude Code draw from the same pool. The $100 plan formerly called Max 5x is now just “Max” and keeps its 5-hour limit.

$200/month10x Pro weeklyNo 5-hour cap (verified)$200 API creditsRenamed Oct 30, 2026

If your billing page suddenly says “Ultra” instead of “Max 20x”, nothing went wrong. Anthropic renamed both Max tiers at the end of October. The rename itself is cosmetic, but the way limits work on the top plan changed, and several other Claude Code limit changes landed this year. This guide covers what Ultra includes, how it compares with every other Claude plan, how the 5-hour and weekly limits work now, how to check your own usage, and when the API is cheaper.

Anthropic does not publish exact message counts for any plan. Every number below is either a list price, a multiplier Anthropic publishes, or our own arithmetic, and we label which is which. For the full cost picture of the coding agent itself, see our Claude Code pricing guide.

What is the Claude Ultra plan?

Ultra is Anthropic’s top individual Claude subscription. According to the Help Center article “What is the Ultra plan?” and the pricing page, it includes:

  • Price: $200 per month for web subscriptions, billed monthly only. Mobile app-store pricing can differ.
  • Usage: 10 times the Pro plan’s usage each week, with one weekly limit across all models plus a separate weekly limit for Fable models. The pricing page says Fable can use up to 50% of the weekly limit.
  • No 5-hour limit for verified accounts. Unverified accounts get a temporary 5-hour limit until Anthropic verifies them.
  • Claude Code in the terminal, VS Code, Cursor and JetBrains IDEs, on the same subscription.
  • $200 a month in Claude API credits, which you can spend on the API but not in Claude Code or the Claude apps.

Existing Max 20x subscribers were moved to Ultra automatically. The official @claudeai post on October 30 said “If you’re a subscriber, there’s nothing you need to do,” and that the 5-hour limit removal was starting “with most existing subscribers” first.

Why Ultra is 10x Pro, not 20x

The old name suggested 20 times Pro, but the pricing page now lists Ultra at “10x Pro usage” per week. Anthropic’s head of product for Claude subscriptions explained the gap on X: when the top plan launched in April 2025 it had only a 5-hour limit, set at 20x Pro. When weekly limits were added later, they were set at 10x Pro, while the 5-hour limit stayed at 20x. The rename, he said, makes the weekly figure the headline. Anthropic’s announcement said weekly limits did not change on October 30, so heavy Max 20x users didn’t lose weekly capacity. They lost a misleading name and, if verified, the 5-hour cap.

Ultra plan is not Ultraplan. Many search results for “claude ultra plan” are about Ultraplan, a Claude Code feature that drafted a plan in a cloud session while your terminal stayed free. It launched as a research preview in the week of April 6–10, 2026 (Claude Code v2.1.92–v2.1.101). Anthropic’s docs now say Ultraplan is no longer available: the /ultraplan command was removed, and the docs point to plan mode or cloud sessions instead. It was never tied to the Ultra subscription.

Claude plans compared: Free, Pro, Max, Ultra, Team, Enterprise

All figures are US list prices from claude.com/pricing and the Help Center on November 7, 2026. “Relative usage” is Anthropic’s own multiplier. Pro is the baseline, and Anthropic doesn’t publish what one “Pro” equals in messages or tokens.

PlanPrice (US)Weekly usage5-hour limitClaude CodeFable modelsAPI credits
Free$0LimitedYesNoNot listedNone
Pro$20/mo, or $17/mo billed annually ($200/yr)1x (baseline)YesYesVia usage credits onlyNone
Max (formerly Max 5x)$100/mo5x ProYes (5x Pro per window)YesUp to half of weekly limit$100/mo
Ultra (formerly Max 20x)$200/mo, monthly only10x ProNo, for verified accountsYesUp to half of weekly limit$200/mo
Team, Standard seat$20/seat/mo annual, $25 monthly“More usage than Pro”Yes (rolling per-seat window)YesNot statedUp to $500/mo pooled per team
Team, Premium seat$100/seat/mo annual, $125 monthly5x a Standard seatYes (rolling per-seat window)YesUp to half of weekly limitPooled with team
Enterprise$20/seat/mo annual, 20-seat minimum, plus usage at API ratesUsage-basedDepends on contractYesNot statedNot stated

Team covers 2 to 150 users. Prices are before tax. Regional checkout can show local currency, VAT and different amounts.

Two details are easy to miss. First, Ultra gives exactly twice Max’s weekly usage (10x vs 5x Pro) for exactly twice the price. Per unit of weekly usage, the two cost the same, and what the extra $100 buys is no 5-hour cap and $100 more in API credits. Second, the included API credits are a separate wallet. The Claude Code help article says monthly API credits on Max, Ultra and Team “can’t be used in Claude Code or the Claude apps.” They’re for your own API projects.

How Claude Code usage limits work now

Every paid individual plan uses the same two-layer system. The difference is which layers apply to you.

How Claude plan limits work: 5-hour window vs weekly pool Pro and Max accounts are capped by two limits at once, a rolling 5-hour session limit and a weekly limit across all models. Verified Ultra accounts have no 5-hour limit and only a weekly limit of 10 times Pro usage, plus a separate weekly limit for Fable models. Claude chat and Claude Code draw from the same pool, and usage credits billed at API rates can continue work after a limit. How Claude plan limits work: 5-hour window vs weekly pool DecodeTheFuture.org Claude Ultra, Claude Max, Claude Pro, Claude Code usage limits, 5-hour limit, weekly limit Two-layer limit model for Claude plans as of November 2026: Pro and Max have a 5-hour and a weekly limit; verified Ultra has only a weekly limit. Diagram image/svg+xml en © DecodeTheFuture.org Which limits stop you? Pro and Max 1. Rolling 5-hour session limit 2. Weekly limit, all models Hit either one and you wait for its reset. Ultra (verified account) 1. 5-hour limit: none 2. Weekly limit = 10x Pro Fable has its own weekly cap (up to 50%). Chat + Claude Code share one pool. After a limit: usage credits at API rates.

Bars are illustrative fill levels, not measured quotas. Source: Anthropic Help Center and pricing page, November 7, 2026.

The 5-hour session limit

Free, Pro and Max accounts, and unverified Ultra accounts, have a rolling 5-hour session limit. It exists to spread load across the day. Anthropic doubled Claude Code’s 5-hour limits for Pro, Max, Team and seat-based Enterprise plans on May 6, 2026 and removed the peak-hours reduction for Pro and Max (our news write-up covers that change). If you hit the 5-hour limit in the middle of a response, Claude Code may finish the current step and show “Usage limit reached · wrapping up”. That small allowance still counts toward your weekly limit and is rolling out gradually.

The weekly limit

Every paid plan has a weekly limit across all models, and Max and Ultra have a second weekly limit for Fable. The weekly limit resets at a fixed time each week assigned to your account, not on a rolling basis, and Settings > Usage shows the next reset. For Claude Code specifically, the Help Center confirms that weekly limits ran 50% higher from May 13 to September 13, 2026 as a promotion. Since September 14 they’ve been permanently 25% above the pre-promotion level, for Pro, Max, Ultra, Team and seat-based Enterprise. Compared with the promotion period, that’s about a 17% reduction (125 vs 150), which is why some users felt their limits shrank in September.

One pool for chat and Claude Code

Anthropic’s help article is explicit: “all activity in both tools counts against the same limits.” A long research chat in the Claude app eats into the same weekly budget as your Claude Code sessions. Your model choice matters too. The pricing page notes that “how fast you reach a limit depends on the model you use,” along with conversation length and file sizes. Opus-heavy work burns the budget faster than Sonnet.

What doesn’t change your limits

  • Message threads (public beta, November 2, 2026): Claude Code now sends only new messages each turn, cutting median time to first token by about 35% on 500K–1M-token requests. Anthropic says token usage stays the same, so this makes Claude Code faster without stretching your plan.
  • Fast mode never counts against plan limits. On Pro, Max and Ultra it’s billed only from usage credits. Fast mode for Opus 4.8 was deprecated on October 30, 2026 and will be removed on November 30, 2026. After that date, switch to Opus 5.5 (fast mode $8/$40 per MTok) or run /fast off.
  • Switching models after a session or weekly limit. Those limits apply across all models, so /model doesn’t get you back in. Only a model-specific limit message, such as “You’ve hit your Opus limit”, lets you continue on another model family.

How to check your Claude usage and Ultra status

  1. In Claude Code, run /usage. According to the Claude Code cost docs, subscribers see plan usage bars and a breakdown of what consumed usage: skills, subagents, plugins and individual MCP servers. You also get recommendations such as trimming a large CLAUDE.md. Press d or w to switch between 24 hours and 7 days, and f (v2.1.305+) to let Claude Code apply the fixes it can. The breakdown only covers this machine, so it misses claude.ai and your other devices.
  2. On the web, open Settings > Usage. It shows your limits, the next weekly reset and your usage-credits spend. On Ultra, if there’s no “Current session” bar, your account is verified and has no 5-hour limit.
  3. Run /status to see what you’re billed against. A Login method row such as “Claude Max account” means subscription. An API key row means Console billing. If ANTHROPIC_API_KEY is set in your environment, Claude Code uses it instead of your subscription and you pay API rates.

Still seeing a 5-hour bar on Ultra? Anthropic verifies accounts automatically from account and payment history, and lifts the temporary limit without any action on your part. Verifying a phone number at checkout does not remove it, per the Ultra help article. New subscribers should expect the 5-hour cap at first.

Should you upgrade to Ultra?

  • Pro, hitting the 5-hour limit a few times a week: try Max first. It raises both the 5-hour and weekly limits fivefold for $100.
  • Max, hitting the 5-hour cap in long sessions but rarely the weekly one: Ultra helps most here. Removing the cap lets you spend your weekly budget in long bursts, which suits multi-hour agent runs.
  • Max, hitting the weekly limit regularly: Ultra doubles the weekly pool. If you’d only run over by a little, usage credits at API rates may cost less than an extra $100 a month.
  • Ultra, still running out: turn on usage credits with a monthly cap, or move automation to an API key. Credits bought on or after December 7, 2026 expire after 12 months.
  • You also build on the API: Ultra’s $200 in monthly API credits lowers the net cost of the subscription, as long as you’d spend that money on the API anyway.
  • Teams: compare Team Premium seats ($100–125 per seat, 5x a Standard seat) instead of individual Ultra subscriptions. Admins get spend controls and pooled API credits.

Ultra vs the API: when is $200 a month worth it?

Claude Code on an API key bills every token. To compare, we reuse a typical agentic coding turn: 200K input tokens, 80% served from cache, plus 20K output tokens, with cache-write costs left out. Prices come from the Claude API pricing page on November 7, 2026. Opus 5.5 and Sonnet 5.5 cache reads now cost 0.05x the input price.

Model (API)Price in / out per MTokCost per turnTurns per month for $100Turns per month for $200
Haiku 5.5 (prompt over 100K)$0.50 / $2.50$0.078~1,280~2,560
Sonnet 5.5$2 / $10$0.296~338~676
Opus 5.5$4 / $20$0.592~169~338

Over 22 working days, $200 of API spend buys about 31 such Sonnet 5.5 turns a day, or about 15 on Opus 5.5. An agent session of a few hours easily runs more turns than that. Anthropic’s own figure, in the Claude Code cost docs, is an average of about $13 per developer per active day and $150–250 per developer per month across enterprise deployments, at API rates.

Our reading: if you use Claude Code most working days, a flat $100–200 plan is almost certainly cheaper than the API at these token profiles. The API wins for irregular use, CI jobs and scripts, and anywhere you need per-token cost control. Ultra makes the most sense for people who’d otherwise spend well above $200 a month in API tokens on Claude Code. That’s what the Max 20x crowd was. Our Sonnet 5.5 pricing guide and Claude models overview cover model choice in more detail.

2026 timeline: Claude Code limit changes

Date (2026)ChangeStatus
May 6Claude Code 5-hour limits doubled (Pro, Max, Team, seat Enterprise); peak-hours reduction removed for Pro and MaxVerified, Anthropic news
May 13 – Sep 13Weekly Claude Code limits +50% (promotion)Verified, Help Center
Sep 14Weekly Claude Code limits permanently +25% vs pre-promotionVerified, Help Center
Oct 30Max 5x → Max, Max 20x → Ultra; 5-hour limit removed for verified Ultra; weekly limits unchangedVerified (Help Center, pricing); date from official @claudeai post
Oct 30Opus 4.8 fast mode deprecated in Claude CodeVerified, Claude Code docs
Nov 2Message threads beta: faster responses, same token usageVerified, Claude blog
Nov 30Opus 4.8 fast mode removedScheduled, Claude Code docs
Dec 7Usage credits bought from this date expire after 12 monthsScheduled, Help Center

What we don’t know yet

  • Exact quotas. Anthropic publishes multipliers, not message or token counts. Real capacity depends on model, context length, attachments and tool use.
  • Verification timing. Anthropic doesn’t say how long it takes to verify a new Ultra account.
  • Annual billing for Max. The pricing page doesn’t list one, and Ultra is explicitly monthly only.
  • When Ultraplan was removed. The docs confirm the removal but not the date, and third-party accounts disagree.

If you’re weighing Claude Code against alternatives, our Cursor vs Claude Code comparison and Fable pricing and limits breakdown cover the other pieces.

Frequently asked questions

What is the Claude Ultra plan?

Ultra is Anthropic’s top individual Claude subscription at $200 per month, formerly called Max 20x. It includes 10 times Pro’s weekly usage, Claude Code, $200 a month in API credits, and no 5-hour limit for verified accounts.

Is Claude Ultra the same as Max 20x?

Yes. Anthropic renamed Max 20x to Ultra on October 30, 2026 and moved existing subscribers over automatically at the same $200 price. Weekly limits did not change. The main difference is that verified Ultra accounts no longer have the 5-hour limit. The $100 Max 5x plan is now simply called Max.

Does Claude Ultra have unlimited usage?

No. Ultra removes the rolling 5-hour limit for verified accounts, but a weekly limit of 10 times Pro usage still applies across all models, plus a separate weekly limit for Fable. Claude chat and Claude Code share that pool.

Why do I still have a 5-hour limit on Ultra?

Your account is probably not verified yet. New Ultra accounts keep a temporary 5-hour limit until Anthropic verifies them from account and payment history, and the limit is then lifted automatically. Verifying a phone number at checkout does not remove it. If Settings > Usage shows no Current session bar, you are verified.

Is Ultraplan part of the Claude Ultra plan?

No. Ultraplan was a Claude Code research-preview feature, launched in April 2026, that drafted plans in a cloud session. Anthropic has since removed it and recommends plan mode or cloud sessions instead. It had nothing to do with the Ultra subscription.

How do I check my Claude Code usage limits?

Run /usage in Claude Code to see plan usage bars and a breakdown by skills, subagents, plugins and MCP servers. On the web, open Settings > Usage to see your limits, the next weekly reset and usage-credits spend. Run /status to confirm whether you are billed to a subscription or an API key.

Can I use Ultra’s $200 API credits in Claude Code?

No. Anthropic says the monthly API credits included with Max, Ultra and Team cannot be used in Claude Code or the Claude apps. They apply to your own Claude API usage. Going past plan limits in Claude Code uses usage credits, billed at standard API rates.

Polish readers can find this guide in Claude Ultra po polsku.

Sources & methodology

  1. Claude Help Center — What is the Ultra plan? (price, 10x weekly usage, verification, Max 20x migration).
  2. Claude — Plans & pricing (Free, Pro, Max, Ultra, Team, Enterprise; API credits; Fable caps).
  3. Claude Help Center — Using Claude Code with your Pro, Max or Ultra plan (shared limits, wrap-up allowance, API credits).
  4. Claude Help Center — What is the Max plan? (5x Pro per 5 hours and per week, weekly reset).
  5. Claude Help Center — What is the Pro plan?
  6. Claude Help Center — Manage usage credits for paid Claude plans.
  7. Claude Help Center — Claude Code May–September 2026 weekly limits promotion and permanent +25% from September 14.
  8. Anthropic — Higher limits for Claude Code (May 6, 2026).
  9. Claude Code docs — Manage costs effectively (/usage, /usage-credits, average cost per developer).
  10. Claude Code docs — Fast mode (Opus 4.8 deprecation and removal dates, pricing).
  11. Claude Code docs — Ultraplan is no longer available.
  12. Claude Code docs — What’s new, Week 15 (April 6–10, 2026): Ultraplan research preview.
  13. Claude blog — Make long-running agents faster with message threads (November 2, 2026).
  14. Claude API Docs — Pricing (Opus 5.5, Sonnet 5.5, Haiku 5.5).
  15. @claudeai on X — “We’re renaming our Max plans” (October 30, 2026).
  16. Let’s Data Science — Anthropic removes five-hour Claude limit for verified Ultra accounts (October 30, 2026; quotes the official announcement).

Plan facts come from Anthropic’s Help Center, pricing page and docs, accessed November 7, 2026. The October 30 date and the 10x-vs-20x explanation come from posts by Anthropic’s official account and its head of product for Claude subscriptions on X. We couldn’t load them directly, but multiple outlets quote them consistently. Per-turn API costs are DecodeTheFuture’s own arithmetic on list prices for an illustrative token profile, not measurements. Anthropic doesn’t publish exact message counts for any plan.

Exit mobile version