Skip to content
AI Tool Verdicts

Claude Pricing: Every Plan, Usage Limit and API Price

Claude pricing for October 2026: Free, Pro, Max, Team and Enterprise prices, how usage limits work, what each plan adds, and API costs per model.

By AI Tool Verdicts Editorial Team · Published

Five blank paper cards in an oak holder, each taller than the last, the tallest in terracotta, illustrating Claude's pricing tiers

Bottom line: Free for light chat, Pro at $20 for most people, Max 20x for heavy daily use. It gives the most usage per dollar of any plan.

Claude pricing starts at $0. As of October 2026, the paid plans are Pro at $20 a month ($17 billed yearly), Max at $100 or $200 a month, and Team at $25 or $125 per seat. Enterprise is sold per seat on yearly contracts, and the API bills per token, from $0.10 per million input tokens on Haiku 5.5.

For most people, Pro is the right first plan. It adds Opus, Claude Code, Research and Claude in Chrome to everything in Free. Move to Max only when you keep hitting the five-hour limit.

How we researched this: Anthropic's pricing page, help center articles and API pricing docs, checked October 28, 2026. The usage-per-dollar math below is our own, based on Anthropic's published usage multiples.

Claude pricing at a glance

All prices are in USD and exclude tax.

PlanMonthlyBilled yearlyUsage per 5-hour sessionWho it's for
Free$0$0BaselineTrying Claude, light chat
Pro$20$17/mo ($200 up front)At least 5x FreeMost individuals
Max 5x$100Not offered5x ProDaily heavy users
Max 20x$200Not offered20x ProAll-day work, agents
Team Standard$25/seat$20/seat1.25x ProTeams of 2 to 150
Team Premium$125/seat$100/seat6.25x ProTeam members who code all day
EnterpriseContact sales$20/seat plus usageBilled by usageLarge companies

Max is monthly only. Anthropic's Max plan article says it's "currently available as a monthly subscription only", and prices in the mobile apps may differ from the web.

Team needs at least two seats and goes up to 150, per the Team plan article. You can mix Standard and Premium seats on one team.

What does each Claude plan include?

Every plan uses the same Claude apps on web, desktop and mobile. The differences are models, a few features, and how much you can use before hitting a limit.

FeatureFreeProMaxTeamEnterprise
Sonnet and HaikuYesYesYesYesYes
Opus 5.5NoYesYesYesYes
Fable 5.1NoUsage credits onlyUp to 50% of weekly limitPremium seats; Standard via creditsDepends on seat type
Web search, memory, Artifacts, connectorsYesYesYesYesYes
Claude CodeNoYesYesYesYes
ResearchNoYesYesYesYes
Claude in ChromeNoYesYesYesYes
Usage credits (pay to keep going)NoYesYesOwner can enableUsage-based
Monthly API creditsNoNo$100 or $200Pooled, up to $500No
SSO, admin and spend controlsNoNoNoYesYes, plus SCIM and audit logs

A few notes on the table:

  • Fable 5.1 is Anthropic's most capable model, and access varies by plan. Per the Fable help article, Max subscribers can spend "up to half their weekly usage limit" on it, and it uses that limit faster than other models. On Pro you pay for it with usage credits.
  • Free got a lot more generous in 2026. Since September 16, Artifacts and Claude's Design, Slides and Docs tools are on every plan, including Free.
  • Max adds priority access at busy times on top of the higher limits.

If you're unsure whether Free is enough, our guide on whether Claude is free covers exactly what the free plan does and doesn't include.

How do Claude usage limits work?

Claude doesn't sell a fixed number of messages. Each plan gets an allowance that resets on a rolling five-hour window, and paid plans also have a weekly limit across all models. You can see both in Settings > Usage.

Three things decide how far your allowance goes:

  1. The model. Opus uses your limit faster than Sonnet, and Fable faster still.
  2. The length of the conversation. Claude rereads the whole chat with every message, so long threads cost more per reply.
  3. Claude Code. It draws from the same pool as chat. Anthropic's usage credits article says "combined usage across both counts toward your limits".

Anthropic doesn't publish limits in tokens or messages. So searches like "how many tokens does Claude Pro have" have no official answer. The only published numbers are multiples: Pro gets at least 5x Free per session, and Max gets 5x or 20x Pro.

What happens when you hit the limit?

On Free, you wait for the window to reset. On Pro and Max, you can turn on usage credits instead. These are prepaid and bill at standard API rates once you pass your plan's limit. They're charged separately from your subscription, you can set a monthly cap, and they generally don't expire.

Which Claude plan gives the most usage per dollar?

This is the comparison Anthropic's pricing page doesn't make. If you divide each plan's price by its published per-session usage, with Pro as 1 unit, almost every plan costs the same per unit. One plan is the exception.

PlanPriceUsage (Pro = 1)Cost per Pro-sized unit
Pro, monthly$201$20
Pro, yearly$171$17
Max 5x$1005$20
Max 20x$20020$10
Team Standard, yearly$201.25$16
Team Premium, yearly$1006.25$16

Max 20x is the best value per unit of usage, at half the price of Pro or Max 5x. Max 5x costs the same per unit as Pro. It's worth buying only if you'd otherwise run out of Pro, not because it's a better deal.

A caveat on this math: the multiples are per five-hour session. Anthropic doesn't say whether the weekly caps scale the same way, so treat the table as a guide, not a guarantee.

Is Claude Pro worth it?

Yes, if you use Claude more than a few times a week. For $20 you get Opus, Claude Code, Research and Claude in Chrome, plus at least five times the free usage per session. If you're already paying for one chatbot, Pro is a reasonable swap; our best AI for coding roundup compares it with ChatGPT and Gemini for coding work.

Pay yearly only if you're sure. $200 up front saves $40 over twelve months. But refunds are limited outside the EU and UK, as covered below.

Is Claude Max worth it?

Max is worth it if Pro's limit stops you most days. Max 5x at $100 gives five times Pro for five times the price, so it buys headroom, not savings. Max 20x at $200 is the better deal per unit and includes $200 a month of API credits.

Those API credits come with fine print. Per Anthropic's API credits article, they cover the API, Managed Agents and the Agent SDK, but not interactive Claude Code or usage credits in the apps. Unused credits expire at the end of each billing cycle.

If you're buying Max mainly for coding, read our Claude Code pricing guide first. It works through when a subscription beats paying per token.

How much does the Claude API cost?

The API bills per million tokens, with no subscription. Prices from Anthropic's API pricing page as of October 2026:

ModelInputCache write (5 min)Cache readOutput
Haiku 5.5 (prompts up to 100K tokens)$0.10$0.125$0.01$0.50
Haiku 5.5 (prompts over 100K tokens)$0.50$0.625$0.05$2.50
Sonnet 5.5$2$2.50$0.10$10
Opus 5.5$4$5$0.20$20
Fable 5.1$10$12.50$0.25$50

Other API costs worth knowing:

  • The Batch API halves input and output prices for jobs that can wait.
  • Opus 5.5, Sonnet 5.5 and Fable 5.1 include their full 1M-token context window at standard prices, with no long-context surcharge.
  • Web search costs $10 per 1,000 searches, plus tokens.
  • Running inference only in the US costs 1.1x.
  • New API accounts get "a small amount of free credits to test the API", but Anthropic doesn't state the amount.

For a sense of scale, a 2,000-word answer is roughly 4,400 output tokens on current models. On Sonnet 5.5 that's under 5 cents of output. Our Claude API pricing guide covers caching, batch discounts and spend limits in detail. Our comparison of Opus 5.5, Fable 5.1 and Sonnet 5.5 helps you pick a model.

Does Claude charge tax, and does price vary by country?

Prices on the pricing page exclude tax. Anthropic's Pro help article says some regions show tax in the displayed price and others add it at checkout, with "pricing in your local currency where supported". It doesn't publish per-country prices.

How do you cancel Claude, and can you get a refund?

On the web or desktop app, go to Settings > Billing > Cancel. On iPhone or Android, cancel through your app store subscription. You keep access until the end of the billing period. Cancel at least 24 hours before renewal to avoid the next charge.

Refunds are the weak spot. Anthropic's policy says payments are non-refundable except where the law or its consumer terms say otherwise. In the EEA and UK you can request a prorated refund within 14 days of purchase. Elsewhere it's case by case through Get help in the app, and requesting a refund doesn't cancel your plan.

Which Claude plan should you choose?

  • Choose Free if you chat with Claude a few times a week and don't need Opus or Claude Code.
  • Choose Pro if you use Claude most days, or want Claude Code to try on real projects.
  • Choose Max 20x if you hit Pro's limit daily. Skip Max 5x unless $200 is out of budget, since it's no cheaper per unit than Pro.
  • Choose Team if two or more people need shared billing and admin controls. Standard seats cost about the same as Pro with 25% more usage.
  • Choose the API if you're building an app or running scripts, and want to pay only for what you use.

FAQ

How much does Claude cost per month?

Claude Free is $0. Pro is $20 a month, or $17 a month billed yearly. Max costs $100 or $200 a month, and Team costs $25 or $125 per seat, as of October 2026.

How much is Claude Pro?

Claude Pro costs $20 a month billed monthly, or $200 a year ($17 a month) billed yearly, plus tax where it applies.

How much is Claude Max?

Max 5x costs $100 a month and Max 20x costs $200 a month. Both are monthly only, with no yearly discount.

Does Claude Pro have a limit?

Yes. Pro has a usage limit that resets every five hours, plus a weekly limit. Anthropic says Pro gets at least five times the per-session usage of Free, but doesn't publish the limit in messages or tokens.

Is there a free trial of Claude Pro?

No. Anthropic's pricing page lists trials only for sales-assisted Enterprise deals. The free plan is the way to try Claude before paying.