Model catalogue · Sentence-Transformers (offene Gewichte)

Mehrsprachiges Embedding

A small, multilingual embedding model for short pieces of text - subject lines, keywords, brief notes. Its very tight context window is the decisive limit: it is not made for whole file passages, but it is made for indexing a mixed-language corpus. It is procured via IONOS in the Berlin data centre; the German location is a property of the service there and not a setting that can be chosen wrongly. The chain for section 203 of the German Criminal Code is nevertheless not closed: the data processing agreement does contain a confidentiality undertaking expressly in awareness of criminal liability under section 203, but the product-specific confirmation for this service is still outstanding.

Usable, but one piece is still open

You can work with it - but part of the safeguard is open: either the provider does not commit to confidentiality itself, or it depends on an agreement you still have to conclude. Which of the two applies to which route is set out in the full matrix.

What it is good for

  • Multilingual at very low resource requirements
  • Available from German data centres, without a US group in the chain
  • The most economical tier for indexing and short-text similarity

Where it stops

  • 128 tokens of context - whole file passages do not fit
  • For a file search across longer documents the larger embedding models in this catalogue are the right choice
  • The product-specific confirmation of the section 203 clause for the AI Model Hub is still outstanding

Status through us

The same model, procured differently, is a different legal position. The status above refers to procurement through us; processing takes place in Germany.

Which route carries it, with which instrument and under which condition, we set out in full to the firm before it commissions us - section 203(4) requires that the chain can be checked, not that it is published. The full matrix, with the sources behind every route, can be requested in the German catalogue after confirming your e-mail address, and we go through it in the conversation.

Is Mehrsprachiges Embedding the right fit?

We go through your tasks and the route that carries them - and say plainly where it does not.

Arrange a conversation