By Jordi Daniels, founder of AI Board · ·
Claude has four paid plans. Pro costs $17 a month billed annually, Max from $100 a month and Team $20 per seat. Enterprise costs $20 per seat plus usage, according to claude.com/pricing on 25 September 2026.
Claude answers in the language you write in, Dutch included. For a small business, Pro or Max fits one decision-maker, Team two to 150 colleagues.
Below, every amount carries its source and check date, followed by what Anthropic itself says about your data and an honest comparison with ChatGPT.
Full disclosure: AI Board runs on Claude itself, so we are not neutral. That is why every statement about Claude links to a page of Anthropic's own.
Which Claude plans are there and what do they cost?
Anthropic sells Claude to individuals and to teams. All prices are per user per month, excluding tax, checked on 25 September 2026. Euro prices in brackets are as shown in the Netherlands.
| Plan | Billed annually | Billed monthly | For whom | Source (checked 2026-09-25) |
|---|---|---|---|---|
| Claude Free | $0 | $0 | one person, to try it out | claude.com/pricing |
| Claude Pro | $17 (€15) | $20 (€18) | one person; Claude Code included | claude.com/pricing |
| Claude Max | not listed | from $100 (€90) | one person; 5 or 20 times the usage of Pro | claude.com/pricing |
| Claude Team, standard seat | $20 (€18) | $25 (€21.04) | 2 to 150 users; SSO and central admin | claude.com/pricing/team |
| Claude Team, premium seat | $100 (€90) | $125 | five times the usage of a standard seat | claude.com/pricing/team |
| Claude Enterprise | $20 plus usage | none | annual billing only; usage cost depends on model and task | claude.com/pricing |
Where only a dollar amount is shown, we could not check the euro price. Three things stand out.
Claude Code is in every paid plan. That is the program Claude uses to work with files on your own computer. Free does not include it.
Pro and Max are personal plans. They fall under Anthropic's consumer terms, even when you use them for work. What that means for your data is two sections down.
Team is the first business plan. You get central billing, admin and sign-in through SSO. According to the Team page, Anthropic does not train on the content of a Team account by default.
Enterprise is for larger organisations. Anthropic lists SCIM, audit logs and configurable retention among its features. The seat price is low, but usage costs come on top.
The comparison with ChatGPT Business, with a worked example for fifteen employees, is in what does AI cost your business.
Which languages does Claude work in?
For the conversation itself, the language you use. Anthropic's help centre says Claude replies in the language you write in. Ask in Dutch, German or French and the answer comes back in that language.
The interface is a different matter. Anthropic lists eleven interface languages for web and desktop, including German, French and Spanish. Dutch is not among them (checked 25 September 2026), so Dutch users see menus in English or another listed language.
Anthropic publishes no quality score for Dutch. Anthropic's language table measures fourteen languages besides English, and Dutch is not one of them. Anthropic adds that Claude is capable in many more languages, and advises you to test.
That test takes ten minutes. Ask Claude to rewrite an existing proposal, email or memo in your language, then judge the tone and the jargon yourself.
The same documentation gives one more tip. State explicitly which language you want the answer in. That is more reliable than counting on Claude to detect it.
What does Anthropic do with your data?
It depends on your plan. Anthropic separates consumer plans (Free, Pro and Max) from commercial plans (Team, Enterprise and the API). Below is only what Anthropic documents itself.
| Free, Pro and Max | Team and Enterprise | |
|---|---|---|
| Terms | Consumer Terms | Commercial Terms |
| Training on your chats | only when the model improvement setting is on | not by default |
| Chats in the app | kept until you delete them, then gone within 30 days | same; Enterprise can set its own retention period |
| With model improvement on | up to 5 years, stripped of identifying data | not applicable |
| Claude Code sessions | 30 days, or 5 years with model improvement on | 30 days as standard |
Sources, checked on 25 September 2026: the privacy centre on retention for consumers and for organisations, and the Claude Code data usage documentation.
For business plans, the commitment is at its firmest. The Commercial Terms bar Anthropic from training on customer content. The privacy centre names one exception: when you send feedback or give consent yourself.
Enterprise can set its own retention periods, according to Anthropic. Qualifying Enterprise customers can also get zero data retention for Claude Code, arranged per organisation.
On Free, Pro and Max, a privacy setting decides. With model improvement on, Anthropic uses new chats for training and keeps them for up to five years. That also applies to Claude Code on such an account.
If you use Pro or Max for work, check how this is set in your privacy settings. Incognito chats are never used for training, according to the privacy centre.
We make no statement about where Anthropic processes your data. If that is a requirement for you, ask Anthropic before you start.
And AI Board? Your company brain and your files stay on your own laptop. With the Claude edition, processing of a question goes through Anthropic. With the local edition, that stays on your machine too.
The Claude subscription is in your name. So the rules in the table apply to whichever plan you choose.
ChatGPT or Claude for your business?
Which one is better depends on your work, not on the brand. Here are five points where they differ or turn out to be the same, each with the vendor's own source.
1. The price per seat is close. In euros as shown in the Netherlands, ChatGPT Business costs €21 per seat per month billed annually, Claude Team €18. Sources: chatgpt.com/pricing and claude.com/pricing, checked 25 September 2026. That gap rarely decides it.
2. Neither trains on business data by default. OpenAI says so for ChatGPT Business and Enterprise, Anthropic for Team and Enterprise. On personal ChatGPT accounts, sharing for model improvement is on by default; on Claude, a privacy setting decides.
3. Team size differs. ChatGPT Business runs from 2 to 200 users, Claude Team from 2 to 150. Beyond that, both mean Enterprise. Claude Enterprise has a public seat price, ChatGPT Enterprise does not.
4. Claude also works on your computer. Claude Code, in every paid plan, works with files in a folder on your computer. AI Board builds on that. If you mainly want a chat window, this matters less.
5. Quality differs by task. There is no winner for everything. A model that is strong at analysis does not automatically write best. Our research on which AI model for which role lists what to test first for each role.
Then test on your own work: one task, two weeks, with both. If your data may not leave the laptop, a local model is a third option. The trade-off between all three is in ChatGPT, Claude or a local model.
Claude on your own company data
Claude in the browser only knows what you give it. You can upload files, or open up mail and documents through a connector. Anthropic lists a Microsoft 365 connection for both Pro and Team.
Claude Code goes a step further. The program runs on your own computer and works with the files in a folder you point it to. Your question and the context it finds go to Anthropic.
Claude Code is the engine of AI Board. We set it up as a personal AI assistant. Your company brain sits as plain files in Obsidian on your laptop, next to your own documents.
You also have mail and calendar connected (Microsoft 365 or Google Workspace). Routines run by themselves, such as a morning update and a weekly review. They run on your laptop, so it needs to be on in the morning.
Your documents do not move to a new platform. Prefer that questions do not go through Anthropic? Then there is a local edition. Local models are weaker than the best cloud models, which shows in heavy analysis.
Why it matters where your files live is in private AI on your own laptop.
For business owners in SMEs, we install this in one day on site: the setup day. What that looks like for your company is something we discuss in a call. You pay for your Claude subscription yourself, directly to Anthropic.
What it is not: a rollout for your whole team. The setup day is for the person who makes the decisions. If fifteen colleagues need to work with Claude, Claude Team is the logical route.
Claude Cowork: tasks Claude carries out itself
Claude Cowork was the mode in which Claude works on tasks by itself, in folders and apps you choose. According to claude.com/product/cowork, it is now simply part of Claude, rolling out to Pro and Max first (checked 25 September 2026).
The idea stays the same. Claude shows at each step which files it opens and which tools it uses. You can schedule tasks, and according to Anthropic they keep running when your laptop is closed.
Frequently asked questions
Is Claude free?
Yes, there is a free version: Claude Free, for one person. Claude Code and more usage start at Pro, for $17 a month billed annually. Source: claude.com/pricing, checked 25 September 2026.
Free also falls under the consumer terms. So set the model improvement privacy setting the way you want it.
What is the difference between Claude Pro and Claude Team?
Pro is a personal plan for one user, under the consumer terms. Team is for 2 to 150 users, with central billing, admin and SSO.
The biggest difference is your data. On Team, Anthropic does not train on your content by default; on Pro, that depends on your privacy setting. Pro costs $17 a month billed annually, a Team seat $20.
Can Claude read my documents?
Yes, if you give it access. Upload files in a chat, or use a connector such as the one for Microsoft 365. Claude Code and Cowork work in folders on your computer that you choose.
Claude does not read your whole drive by itself. Connecting everything your business knows in good order is work. That is exactly what we do on the setup day.
Is Claude safe for company data?
On the business plans, Anthropic does not train on your content by default, and the retention periods are documented. Beyond that, safety depends on your own rules: which tool is allowed, and which data does not belong in it.
Put that in a short policy. An example you can adopt is in AI policy example. How AI Board handles your data, and what we deliberately do not claim, is on our security page.