Claude’s plans used to be pretty straightforward, but as more features have been added, the differences between subscriptions have become harder to keep track of. Choosing the right plan isn’t as simple as it once was. In this article, we’ll take a closer look at every Claude plan, breaking down its features, pricing, and usage limits to help you decide which one offers the best fit for your needs.
A Claude AI subscription costs $20 per month for the Pro plan and $100 or $200 per month for the Max plan. Claude also has a free plan.
The table below compares all plans Claude currently offers at a glance. Three plans are for individuals and two plans are for teams.
| Plan | Monthly billing | Annual billing | Usage |
|---|---|---|---|
| Free | $0 | $0 | Base limits |
| Pro | $20 per month | $17 per month | At least 5x Free |
| Max 5x | $100 per month | — | 5x Pro |
| Max 20x | $200 per month | — | 20x Pro |
| Team Standard | $25 per seat | $20 per seat | More than Pro |
| Team Premium | $125 per seat | $100 per seat | 5x Team Standard |
| Enterprise | — | $20 per seat | Paid separately at API rates |
The prices listed above exclude taxes. In some countries, Claude charges in your local currency, so the final price may vary slightly depending on exchange rates and local pricing.
Claude plans differ in usage limits, available models and extra features.
Yes, Claude has a free plan. It includes:
The free plan uses Sonnet 5.5 by default. Free users can also switch to Haiku 5.5 and to the older models Sonnet 5, Sonnet 4.6 and Haiku 4.5. Opus 5.5, Fable 5.1, Claude Code and Projects require a paid plan.
The free plan has no fixed number of messages. The number of messages you can send depends on the length of the conversation and the model, and the limit resets every five hours. The usage page in Settings is not available on the free plan.
The cheapest Claude paid plan is called Claude Pro, and it costs $20 per month with monthly billing, while with annual billing, it costs $200 per year, which equals $17 per month. So, when paying annually, you're getting a 17% discount on the plan.
In addition to everything included in Free, Pro includes at least five times more usage per session and gives access to the Opus models, including Opus 5.5. Pro also includes:
Note that Claude's best frontier model, Fable 5.1, is not included in Pro — instead, it is paid separately with usage credits at API rates: $10 per million input tokens and $50 per million output tokens. Pro also doesn't include access to the Claude API — the API is billed separately.
You might be wondering — is Pro enough for heavy usage?
For writing and light coding, yes. Even with frequent use, you’re unlikely to run into usage limits too often. However, agentic tasks, in-depth research, and computer use can eat through your allowance much faster, potentially leaving you with only a few hours of work before hitting a limit.
What makes things more frustrating is Claude’s multi-layered usage limits. Your usage resets on separate five-hour and weekly schedules, so even if you still have some of your weekly allowance left, you may find yourself waiting for the five-hour limit to reset before you can continue working.
For heavy users, Claude offers a more expensive plan with advanced features and extended usage limits — Claude Max. If you find yourself running out of usage often, this plan might be for you.
This plan has two tiers: Max 5x costs $100 per month and gives five times more usage per session than Pro, and Max 20x costs $200 per month and gives twenty times more. Both tiers are billed monthly only.
In addition to higher limits, Max includes:
The API credits can be used on the Claude Platform for the Claude API, Managed Agents and the Agent SDK. To use them, you need to link a Console account. The credits do not apply to Claude chat or Claude Code, and unused credits expire at the end of each billing cycle.
The Team plan is for teams of 2 to 150 people and is paid per seat. There are two seat types, and a team can combine them. A Standard seat gives more usage than Pro. A Premium seat gives five times more usage than a Standard seat and includes Fable 5.1, as on Max.
Team includes all Pro features and adds admin tools:
Claude Enterprise is for organizations with 20 or more users. It costs $20 per seat per month, billed annually, plus usage at API rates. Enterprise includes all Team features and adds compliance tools, such as audit logs and custom data retention. A HIPAA-ready version is also available.
Anthropic also offers an Education plan for universities. It gives students and staff discounted access, and its terms are agreed with Anthropic's sales team.
All Claude plans have usage limits. Usage is counted in sessions that reset every five hours, and paid plans also have a weekly limit.
Chats on the web, desktop and mobile apps and work in Claude Code all count toward the same limit.
Claude limits are not measured in messages. Long conversations and larger models use more of the limit, so a session with Opus 5.5 ends sooner than a session with Sonnet 5.5.
When you reach the limit, you can:
Usage credits are billed separately from the subscription, and you can set a monthly spending limit for them. Pro and Max users can check their session and weekly usage in Settings > Usage.
The Claude API lets developers use Claude models in their own apps and pay for each request. API billing is separate from Claude subscriptions. API prices are set per million tokens, and one token is a piece of text about three-quarters of an English word long.
The table below shows API prices for Anthropic's current models:
| Model | Input per 1M tokens | Output per 1M tokens |
|---|---|---|
| Fable 5.1 | $10 | $50 |
| Opus 5.5 | $4 | $20 |
| Sonnet 5.5 | $2 | $10 |
| Haiku 5.5 | $0.10 | $0.50 |
For prompts longer than 100,000 tokens, Haiku 5.5 costs $0.50 per million input tokens and $2.50 per million output tokens.
For example, a request with 2,000 input tokens and 1,000 output tokens costs about $0.03 on Opus 5.5 and about $0.01 on Sonnet 5.5.
Batch processing costs 50% less than standard requests, and fast mode for Opus 5.5 costs twice the standard price. Web search costs an extra $10 per 1,000 searches.
On the API, you pay for every token, so the cost of a task depends on how much text Claude reads and writes. Opus 5.5 and Sonnet 5.5 always use extended thinking, and thinking tokens are billed as output tokens. In a chat, every new message sends the whole conversation to the model again, so long chats and agentic tasks cost more than a single request.
The table below shows how much typical tasks cost at API rates, based on our own work with Claude:
| Task | Sonnet 5.5 | Opus 5.5 |
|---|---|---|
| Writing a blog article with two or three rounds of edits | about $0.20 | about $0.40 |
| In-depth research with about 25 web searches | about $0.75 | about $1.30 |
| Building a landing page in Claude Code over several iterations | about $1.50 | about $3 |
For example, writing an article with a few rounds of edits takes about 30,000 input tokens and 15,000 output tokens. In-depth research takes about 200,000 input tokens, because Claude reads every page it finds, and the web searches are billed on top. Building a landing page in Claude Code can take around 3 million tokens, but most of them are cache reads, which cost a twentieth of the normal input price.
For writing and research, the API costs about the same as a subscription. If you write an article every working day and run two research tasks a week, the API costs about $18 per month on Opus 5.5 and about $10 per month on Sonnet 5.5. Claude Pro costs $20 per month and also includes the Claude apps and Claude Code.
For agentic coding, a subscription is much cheaper. On the API, Claude Code costs about $13 per developer per active day on average, or $150–250 per month. A developer who uses Claude Code every day pays less on Pro at $20 or on Max at $100–200 per month, as long as the work fits within the plan's limits.
If you want to use Claude a lot for work or personal projects, then Claude Pro is certainly worth it, because without it you'll be stuck waiting for limits to reset way too often.
The plan costs $20 per month, the same as ChatGPT Plus and Google AI Pro, and includes Opus 5.5 and Claude Code — some of the best coding tools in the world as of this writing. With annual billing the price is even lower at $17 per month.
But if you only use Claude a few times a week, or for odd questions during the day, chances are you'll be set with the free plan. It includes Sonnet 5.5, a very capable everyday model.
Finally, Claude Max is only worth it if you've tried Pro and found that you exhaust the usage limits regularly, and that optimization strategies don't help. For example, when you work in Claude Code for several hours a day. Max mainly adds usage. It also includes Fable 5.1 and API credits, which are useful only if you build on the Claude Platform.
A Claude subscription is not a good fit if you need images or video, because Claude can't generate images on any plan. It is also an expensive way to use several AI models, since each company charges separately: Claude Pro and ChatGPT Plus together cost $40 per month.
Claude, ChatGPT and Gemini have similar prices. The main plan costs about $20 per month, and the plans for heavy users cost $100 and $200 per month. ChatGPT also has a cheaper Go plan for $8 per month.
| Price per month | Claude | ChatGPT | Gemini |
|---|---|---|---|
| $0 | Free | Free | Free |
| $8 | — | Go | — |
| About $20 | Pro | Plus | Google AI Pro |
| $100 | Max 5x | Pro | Google AI Ultra |
| $200 | Max 20x | Pro | Google AI Ultra |
ChatGPT and Gemini can generate images in the chat, and Google AI Pro includes 5 TB of cloud storage. Claude cannot generate images, but its paid plans include Claude Code.
Overchat AI is an all-in-one AI app we've built that gives access to Claude and to models from OpenAI, Google, xAI and other companies with one subscription. Overchat AI has the following Claude models:
The same subscription also includes other chat models, such as GPT-6 and Gemini 4 Argon, and image and video models, such as Nano Banana 2 and Seedance 2.5.
Overchat AI charges credits for each message, and the number of credits depends on the model. The Pro plan includes 8,000 credits per month for $14.99. The table below shows how many credits each Claude model uses per message:
| Model | Credits per message |
|---|---|
| Claude Sonnet 4.6 | 2 |
| Claude 5 Sonnet | 4 |
| Claude Opus 5.5 | 110 |
| Claude Fable 5.1 | 220 |
For example, we asked Claude Opus 5.5 in Overchat AI to write a short email to a client about moving a website launch to next week. The reply cost 110 credits, which is about $0.21 on the monthly plan.
In Overchat AI, Fable 5.1 is available on the $14.99 plan. Among Claude's individual plans, Fable 5.1 is included only in Max, which costs from $100 per month, and on Pro it is paid separately. Overchat AI is also useful for people who use Claude together with ChatGPT or Gemini, or who need images and video, because one plan covers all of these models.
Since Overchat AI is an AI aggregator, you can access the same models through it, but you can't access Claude's proprietary features like Claude Code, Projects or Claude's connectors, so if those are important for your work, you might be better served sticking to the first-party app. On the other hand, if cost efficiency and access to multiple models in one place are more important for you, then it may make sense to consider an AI aggregator such as Overchat AI.
Below are answers to common questions about Claude subscriptions.
Yes. Claude Pro costs $20 per month with monthly billing, or $200 per year with annual billing, which equals $17 per month. Claude Max costs $100 or $200 per month.
Yes. Claude Code is included in all paid Claude plans, starting with Pro. Claude Code uses the same usage limits as chat.
No. The Pro plan does not include the Claude API. Max and Team plans include monthly API credits for the Claude Platform: $100 on Max 5x, $200 on Max 20x and up to $500 on Team. API usage beyond these credits is billed per token.
Claude Pro is worth it without coding if you use Claude every day for writing, research or working with documents. Pro includes Opus 5.5, Projects and at least five times the free usage. For a few questions a week, the free plan is enough.
Claude Pro gives at least five times the free plan's usage per five-hour session, plus a weekly limit. This is enough for regular daily chats. Long Claude Code sessions and Opus 5.5 use the limit faster. If you often reach the limit, you can turn on usage credits or upgrade to Max.
Claude Pro and ChatGPT Plus both cost $20 per month, and both companies charge $100 and $200 per month for their top plans. Claude Pro costs $17 per month with annual billing. ChatGPT also has a cheaper Go plan for $8 per month.
Claude is the stronger choice for long writing and coding. ChatGPT has more built-in tools, including image generation in the chat. Both chatbots are compared in detail in Claude vs ChatGPT.
Claude payments are generally non-refundable, unless the Consumer Terms or local law say otherwise. To check if you are eligible, open Get help in the Claude menu and choose Claude Refund Request. In the European Economic Area and the UK, you can request a refund in the app within the 14-day withdrawal period. If you subscribed through the App Store, the refund is handled by Apple.
