Model catalogue · DeepL SE
DeepL
DeepL translates, it does not write - in cross-border matters it is often the first AI tool in use and usually also the first where the confidentiality question becomes practical, because a pleading to be translated regularly contains client data. The German vendor processes inside the EU and commits, in the paid tiers, neither to store texts nor to use them for training; that is not yet a closed chain. Because a confidentiality agreement under section 203 of the German Criminal Code is not part of the standard offering when buying directly - it would have to be agreed separately. The free tier is out of the question for client texts in any case.
In the catalogue because you need to know this as well.
What it is good for
- German vendor with EU processing - the third-country question does not arise
- Mentioned positively in the German Federal Bar's environment; that applies to no other specific product
- Narrow task cut: translation instead of free text generation, and correspondingly low hallucination risk
Where it stops
- No confidentiality agreement under section 203 of the German Criminal Code in the standard offering - it has to be agreed separately
- The free tier is not suitable for client texts
- No substitute for a language model: no summarising, no drafting, no analysis
Routes examined
The same model, procured differently, is a different legal position. What is examined per route is the contract chain, not the model.
| Route | Section 203 status | Version on this route |
|---|---|---|
| Direct purchase from the vendor (business customers) | conditional | - |
The conditions behind a conditional status - which piece is open on which route and who closes it - are set out in the full matrix in the German catalogue, and we go through them in the conversation. The German page for DeepL carries the sources with their retrieval dates.
Is DeepL the right fit?
We go through your tasks and the route that carries them - and say plainly where it does not.
Arrange a conversation