Photo on Pexels
Three consumer tiers and a few extras
Anthropic sells Claude to individuals through three tiers: Free, Pro, and Max. Max comes in two sizes, one with 5x Pro's usage and one with 20x. On top of that sit API credits that come bundled with the larger plans, and a Team option for groups. The current list lives on Anthropic's pricing page, and the facts in this article come from that page and from Anthropic's October 7, 2026 announcement of Claude Haiku 5.5.
The ladder is simpler than many software price lists because the main thing that changes from step to step is how much you can use Claude before you hit a limit. Price roughly follows that. Free costs nothing, Pro costs about what a streaming bundle costs, and Max costs five times that or more. What you need to figure out is where your own habits sit on that scale.
This guide covers each tier, puts the numbers in one table, explains the API credits that arrived with Max, and then matches the plans to four kinds of people: a student, a freelancer, a small business owner, and a heavy developer. It also compares the overall shape of Claude's lineup with Google's, which we covered in detail in our Google AI plans guide. If you have never tried an assistant at all, our list of free AI chatbots in 2026 is a gentler place to begin.
A caution before the details. Plans and limits can change, and I only state what Anthropic has published. Where something is unclear, I say so and point you to the source. Check claude.com/pricing before you pay for anything.
What the Free plan is good for
Claude Free costs $0. Anthropic's pricing page is the place to read the exact list of what it includes today, and I will not copy a feature list here that might be out of date by the time you read this. What matters for planning is how people tend to use a free tier: as a trial, as a light daily helper, or as a backup when a paid account runs out of room.
As a trial, Free lets you find out whether Claude's writing style and way of answering suit you. Models differ in tone. Some people like Claude's measured, longer replies, and others want something punchier. A week of real tasks tells you more than any review, including this one. Use it on work you actually need to do, such as rewriting a tricky email or summarizing a long document, and compare the output with what you would have written yourself.
As a light daily helper, Free works for people who ask a handful of questions a day. If you never see a message telling you that you have reached a limit, you have no reason to pay. That sounds obvious, yet many people subscribe on day one and then use the assistant twice a week.
The sign that you have outgrown Free is simple. You get stopped mid-task, you wait, and the wait costs you something. When that happens a few times in one week, Pro starts to look sensible. Until then, keep your $20.
Claude Pro at $20 per month
Claude Pro costs $20 per month when billed monthly, or $17 per month when billed annually, with $200 paid up front for the year. The annual route saves $40 compared with twelve monthly payments of $20. It also commits you for twelve months, so it suits people who already know they will keep using Claude.
The practical reason to buy Pro is more room. Pro is the unit that Anthropic uses to describe the bigger plans, since Max is defined as 5x or 20x Pro's usage. That makes Pro the reference point for the whole lineup. If you want to know whether a bigger plan would help you, the question is whether Pro's allowance runs out on you.
I have not listed which model each consumer plan includes, because Anthropic's public material does not give me a verified answer to cite. The lineup is Opus 5.5, Sonnet 5.5, and Haiku 5.5, and the plan page shows what your tier can reach. If the strongest model is the reason you are shopping, read that page closely and confirm it before you pay. Our article on Claude Opus 5.5 explains what that model does well, which helps you judge whether you need it.
Pro is the plan most individuals should consider first. It costs the same as Google's mid-priced tier, and the step up from Free is clear enough to test over a single month. Paying monthly first and switching to annual later is a sensible way to limit risk.
Claude Max from $100 per month
Claude Max starts at $100 per month and comes in two sizes: 5x Pro's usage or 20x Pro's usage. The pricing page lists no separate annual price for Max, so I cannot give you an annual saving the way I can for Pro. I also will not state a dollar price for the 20x size here. Open the pricing page to see the exact monthly figure for the tier you want.
Max exists for people who work with Claude for hours at a stretch. A person who keeps a long coding session open all day, or who feeds the assistant large documents again and again, can burn through a Pro allowance quickly. For that person, five times the room changes the workday. For a casual user, it is an expensive way to buy capacity that sits unused.
The multipliers are measured against Pro, so the logic is consistent. If Pro lasts you most of the month but runs out in the last week, the 5x size probably covers you with plenty to spare. The 20x size is for people who would run into the 5x ceiling as well, and that is a small group. Start with the lower size and move up only if you hit its limit.
Max also comes with monthly API credits, which we cover in their own section below. Those credits matter if you build anything with Claude's API, because they turn part of the subscription price into spending money for your projects.
Side-by-side comparison
The table below collects the verified facts for each tier. Where Anthropic's public pages give me no confirmed detail, I say to check the page rather than guess.
| Feature | Free | Pro | Max 5x | Max 20x |
|---|---|---|---|---|
| Monthly price | $0 | $20 billed monthly | From $100 | Check the pricing page |
| Annual option | Not applicable | $17 per month, $200 paid up front | None listed | None listed |
| Usage | Baseline | Reference level | 5x Pro | 20x Pro |
| Monthly API credits | Not listed | Not listed | $100 | $200 |
| Model access | Check the plan page | Check the plan page | Check the plan page | Check the plan page |
| Best fit | Trying Claude, light use | Regular daily use | Long daily sessions | Very heavy daily use |
Where the table says Not listed, the sources I used did not mention that item for the tier. That is different from saying the feature is missing, so confirm on Anthropic's page if a row decides your purchase. The Team option is not in the table because its price is not one I can verify, and it is covered separately below.
What the API credits mean
Anthropic's Haiku 5.5 announcement added a benefit to the bigger plans. Max 5x subscribers receive $100 in monthly API credits, and Max 20x subscribers receive $200. Those figures match the starting price of Max 5x, so the credit offsets a large share of the subscription for anyone who would be buying API access anyway.
To see why that matters, you need to know how the API is billed. The consumer apps charge a flat monthly fee for a usage allowance. The API charges per token, which is a small chunk of text, and the bill grows with how much you send and receive. Anthropic prices Haiku 5.5, its cheapest and fastest small model, at $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100,000 tokens. At those rates, $100 of credit goes a long way for light automation, such as classifying support messages or summarizing a batch of notes.
The same announcement says Sonnet 5.5 cache reads were cut by 50%. Cache reads apply when you send the same long prefix, such as a big instruction block or a document, over and over, so a cheaper read helps apps that reuse context heavily. If you are curious how Haiku fits into the family, our Haiku 5.5 explainer covers it, and our API pricing comparison lines up per-token rates across providers.
One point is easy to miss. API credits only have value if you use the API. A person who chats in the app and never writes code gets nothing from them, so do not let the credit sway you into Max unless you have a project that will spend it.
The team angle
Teams have their own option. The Haiku 5.5 announcement says Team receives up to $500 in API credits pooled across users. Pooled means the credits belong to the group, so one person who builds a heavy tool can draw on the shared balance while others use little. I do not have a verified Team price to quote, so ask Anthropic or check the pricing page for the current figure and seat rules.
For a small group, the choice is between giving everyone their own individual plan and using a Team arrangement. Individual plans are simple. Each person pays, each person has their own usage, and nobody depends on anyone else. A shared setup makes more sense when several people need the same tools and one person manages billing.
A team where only one or two people use Claude heavily might do well to buy Pro for most members and a Max plan for the heavy users, if that mix is allowed under the terms. Check that on the pricing page rather than assuming it. Before you move a whole department, run a pilot with three or four people for a month and see who actually uses it.
Privacy and data handling also differ between consumer plans and business plans in most AI products. If you will paste client documents or internal files, read Anthropic's terms for the plan you pick. That reading takes ten minutes and can save you a difficult conversation later.
Which plan fits a student
A student should begin on Free and stay there until limits get in the way. Coursework help rarely needs enormous capacity: explaining a concept, outlining an essay, checking a proof, or quizzing yourself before an exam. A few dozen questions in an evening is a normal load, and Free handles loads like that for most people.
Pro becomes reasonable during crunch periods. If you are writing a thesis, reading dense papers, or working through a large coding project, you may find yourself stopped at the worst moment. A single month of Pro around finals can cost less than a textbook, and you can cancel afterward. The annual price of $200 rarely suits a student because the need comes in bursts.
Max is not a student plan. Starting at $100 per month, it costs more than most people spend on groceries in a week, and a student rarely produces enough daily work to fill it. If you find yourself wanting it, you are likely doing professional-grade work, and a different budget should cover it.
One habit helps at any tier. Ask Claude to explain its reasoning, then check the answer against your textbook or a teacher. The assistant is a good tutor and an unreliable authority, and it will state wrong things with confidence now and then.
Which plan fits a freelancer
Most freelancers should buy Pro. Your time has a price, and a tool that saves an hour a week pays back $20 quickly. Writers use it for outlines and first drafts, designers for client copy, consultants for summarizing meetings, and analysts for explaining numbers in plain language.
The annual option deserves a look once you have a few months of habit behind you. At $17 per month, paid as $200 up front, you save $40 a year, and the cost is a commitment. Freelance income comes in lumps, so some people prefer the monthly plan and its flexibility even at a higher price.
Move up to Max when waiting costs you money. If you run into your limit in the middle of a client deadline, the loss is greater than the price gap. Ask yourself how many times it happened in the last month. Once or twice, plan your work around it. Every week, upgrade.
Tax treatment depends on where you live, so ask an accountant whether the subscription counts as a business expense for you. Keep the receipts in any case.
Which plan fits a small business owner
A small business owner should treat the decision as a question about time saved. Product descriptions, customer replies, meeting notes, policy drafts, and ad copy all sit comfortably within Pro for a typical owner. Start with yourself on Pro, track which tasks you hand to Claude for a month, and then decide whether anyone else needs access.
Buying for staff raises a different set of questions. Who needs it daily, who will try it once and forget, and who handles customer data that should stay out of the tool? Answer those before you buy seats. A simple written rule about what may and may not be pasted into the assistant prevents most problems.
If your business builds a customer-facing tool on Claude, such as a chat helper on your website, the API matters more than the app plan. The credits that come with Max or Team can fund early experiments, but a live product will have its own usage bill. Our ChatGPT, Claude, and Gemini comparison can help you decide whether Claude is the right assistant before you commit.
Max is rarely right for a business owner who uses the chat app alone. It makes sense when the owner also runs coding or automation work that eats credits and allowance every day.
Which plan fits a heavy developer
A heavy developer is the person Max was built for. Long coding sessions, large codebases pasted into context, and repeated test-and-fix loops use a lot of allowance. If Pro leaves you waiting regularly, the 5x size at $100 per month gives you room and returns $100 in API credits for your own projects. That credit can pay for scripts, evaluations, or a small internal tool.
The 20x size is for the person who would exhaust even that. It comes with $200 in monthly API credits, and its price is on the pricing page. Think about whether your work is truly five times heavier than a Pro user's before choosing the larger size, because the move from 5x to 20x costs real money and many people never use the extra.
Developers who ship products should also look at cost per token. Haiku 5.5 at $0.10 input and $0.50 output per million tokens, up to 100,000-token prompts, is the cheap option for high-volume tasks. For harder reasoning, a larger model costs more per token but may finish the job in fewer attempts. The comparison article linked earlier helps you do that arithmetic.
How to decide in four questions
Four questions settle most cases. First, do you hit limits on Free? If you do not, stay on Free. Second, does your work need the stronger models? Check the plan page for which tier reaches them. Third, do you lose paid time when a limit hits? That points toward Pro or higher. Fourth, will you actually spend API credits? If you will not, ignore them when comparing Max with Pro.
After the questions, run an experiment. Pay for one month at the lowest tier that might work and keep a rough tally of how often you hit a limit. A note on your phone is enough. Subscriptions are easy to start and easy to forget, so set a calendar reminder for day 28 to review.
If you never touched the ceiling, you may be able to step down. If you hit it daily, step up one tier and repeat the experiment. This approach costs a few dollars more than guessing right on the first try, and it saves you from long commitments to the wrong plan.
Resist the urge to buy the biggest plan so you never think about limits again. The premium for that peace of mind is large, and a month of data will show you what you really need.
How this compares with Google's plans
At a high level, the two companies build their lineups in a similar way. Both offer a free tier, a mid-priced tier close to $20 per month, and a high-usage tier starting near $100. Google's mid tier is AI Pro at $19.99 and its top tier starts at $99.99 for AI Ultra, which our Google AI plans guide walks through. Claude's equivalents are Pro at $20 and Max from $100.
The differences show up around the edges. Google adds a cheaper Plus tier at $4.99 and bundles large amounts of storage and creative credits, so its plans appeal to people who want a package that includes a lot besides chat. Claude's public plan structure is leaner, with usage multipliers as the main variable, and its API credits on Max speak to people who build things.
Which is better depends on the work. If you want storage, video tools, and a cheap entry step, Google's bundle has more to offer. If your day is writing, analysis, and code, and you prefer a simple ladder where the only question is how much you use, Claude's lineup is easier to reason about. Many people try the free tiers of both and keep whichever one they open first each morning. Our comparison of ChatGPT, Claude, and Gemini goes into how the assistants themselves differ.
Prices and limits can change
Every figure here comes from Anthropic's published pages as of October 2026. Plans and limits can change, sometimes with little notice, and a promotion or restructure could make this article stale. Treat it as a map of how the tiers relate to each other, and let claude.com/pricing have the last word on the number you pay.
Regional pricing, local taxes, and currency can also change what you see at checkout. If you live outside the United States, compare your local price with the dollar figures here before you decide. Billing terms for annual plans, such as refunds and renewals, are worth reading in full, since a $200 payment up front is harder to undo than a monthly one.
Finally, usage limits are the least visible part of any plan. Anthropic describes Max as 5x or 20x Pro's usage, and the exact amounts can shift as models and demand change. That is why the one-month experiment beats any chart. Your own tally of how often you hit the ceiling is the most reliable guide you will find.
Frequently asked questions
How much does each Claude plan cost?
Free costs $0. Pro is $20 per month billed monthly, or $17 per month billed annually with $200 paid up front. Max starts at $100 per month and offers 5x or 20x Pro's usage. The pricing page lists no separate annual Max price, so check the page for the exact figure on the 20x tier.
What is the difference between Max 5x and Max 20x?
Max 5x gives you five times Pro's usage and Max 20x gives you twenty times Pro's usage. Anthropic's October 7, 2026 Haiku 5.5 post also says Max 5x subscribers receive $100 in monthly API credits and Max 20x subscribers receive $200.
Which Claude models come with each plan?
Anthropic's lineup is Opus 5.5, Sonnet 5.5, and Haiku 5.5, but which model each consumer plan includes can change. Open the plan page on claude.com/pricing to see the current model access for the tier you are considering.
Do Claude subscriptions include API credits?
Some do. Per Anthropic, Max 5x subscribers receive $100 in monthly API credits, Max 20x subscribers receive $200, and Team receives up to $500 pooled across users. The announcement does not list API credits for Free or Pro.
Is Claude Pro or Max better for a freelancer?
Most freelancers should start on Pro at $20 per month and move to Max only after hitting usage limits repeatedly. Max starts at $100 per month, so it makes sense when waiting for limits to reset costs you paid work.
How do Claude plans compare with Google's AI plans?
Both companies use a free tier, a mid-priced tier near $20, and a high-usage tier from about $100. Google adds a $4.99 Plus tier and bundles storage, while Claude lists Free, Pro, and Max. Our Google AI plans guide has the full Google breakdown.