Claude AI explained: what is it and why is it different?

In the Netherlands, ChatGPT has almost become a generic name, while for many people Claude is a name they may have seen in passing at most. That's a shame, because if you write in Dutch, Claude is probably the model that takes the most work off your hands.

Claude is the AI assistant from Anthropic, an American AI company that has positioned itself from the start on the safety and reliability of its models. In practice you notice that in the character of the answers: Claude is less quick to guess, more likely to say when something is uncertain, and writes more calmly than the competition. This page explains which models there are, what it costs, where Claude excels and where it lets you down.

Which Claude models are there?

Anthropic runs three models side by side. They all have a context window of a million tokens: that's the amount of text the model can keep in view at once, including your attachments and the whole conversation so far. As a rule of thumb, a thousand tokens is about 750 words, so a million tokens comes down to several thousand pages.

Sonnet 5 is the model you'll use most: fast enough for a conversation, strong enough for serious work. It's also the model in the free version of Claude, and that's notable. Where other providers give their free users a cut-down model, with Claude you get the same model paying users work with all day for free — just with fewer messages.

Opus 5 is the reasoning model. Reasoning means the model first works out step by step how to reach an answer before it responds. That takes longer, but with a complicated legal text, a tricky bug or a multi-step calculation it makes a noticeable difference.

Fable 5 is the top model and, at $10 input and $50 output per million tokens, also the most expensive model on the market. Important detail: Fable 5 is not included in Claude Pro. In the apps you pay for it with prepaid credits.

Why Claude writes the best Dutch

This is the argument that matters for Dutch users. In our own assessment of Dutch language quality, Claude is the only provider with a 5 out of 5.

The difference isn't in spelling mistakes — none of the models make those. It's in what you still have to do afterwards. With ChatGPT you usually rewrite a few sentences because they seem to come straight from English. With Claude you can more often just use a text as is. Ask for an informal tone and it stays informal to the end, instead of falling back to businesslike Dutch after two paragraphs. The full comparison is in which AI speaks the best Dutch.

What's included in Claude Pro?

Claude Pro includes two things you won't find at the competition in the same price range.

Claude Code is a programming environment in which Claude works independently in your project folder: reading files, writing code, running it and correcting itself based on the error messages. For developers this is the reason to get Claude Pro; how it compares with Codex and GitHub Copilot is covered in coding with AI.

Claude Cowork is included as well. Together with the context window of a million tokens, that makes Claude Pro suited to work that involves lots of documents.

There's one more reason Claude stands out among Dutch business users: on the paid consumer plans, training on your conversations is off by default. If you turn it on yourself, your data is kept longer. That's the reverse of the default you get with many other services.

How much does Claude cost?

Anthropic bills in dollars, excluding VAT. At an exchange rate of $1 = €0.8687 and 21% Dutch VAT, you end up with these amounts.

That puts Claude Pro in exactly the same price range as ChatGPT Plus and Google AI Pro. Which of the three gives you the most depends on your work; we make that comparison in ChatGPT Plus vs Claude Pro vs Google AI Pro.

Where Claude falls short

Claude isn't a jack of all trades, and on a few points it clearly loses.

No images, no video, no voice generation. ChatGPT Plus gives you GPT Image 2 and Sora 2 for the same price, and Google has Nano Banana Pro and Veo 3.1. With Claude you get text and code, and nothing else. If you regularly need images, Claude can't be your only subscription.

The top model isn't included. You have to buy Fable 5 separately with credits. So you pay €21.02 and still don't have the provider's best model.

The usage windows fill up faster than you think. Claude works with windows instead of a daily counter. Upload a few large documents in a row and you'll hit your limit by mid-afternoon. With text chat you barely notice, with document work all the more.

American company, American processing. If the question "where is my data stored" is decisive, Claude isn't the simplest answer. You're more likely to end up with Mistral, which processes in France and Sweden.

Less of an ecosystem. There's no office suite around it, no storage, no search engine. You get a chat, and that chat is good.

Who is Claude the right choice for?

In short: get Claude if you mainly work with text and code, not if you're looking for a broad package.

• You write a lot of Dutch — Claude is the best choice, by a wide margin.
• You program — Claude Code in Pro is one of the most affordable ways to develop seriously with AI.
• You work with long documents and don't want training on your data — the context window of a million tokens plus training switched off by default fit that well.
• You want images, video or an Office integration — then Claude isn't enough; take a look at our overview of ChatGPT alternatives.
• You use Claude alongside other models — then you pay €21 two or three times over, and a bundle subscription is cheaper.

If you want to know exactly how Claude compares with GPT-5.6 and Gemini per task, you'll find that in GPT-5.6 vs Claude 5 vs Gemini 3.5. An overview of every model from every provider is in every AI model at a glance.

Conclusion