Model catalogue · Cohere
Cohere Embedding
Cohere's embedding model arranges files, pleadings and correspondence for search and is expressly multilingual: a German question also finds the English or French passage. Like every embedding model it prepares search, not assessment. Procurement runs via Amazon Bedrock - region-bound in Ireland or through the European inference profile, without operator access and without storage in the delivered state. For section 203 of the German Criminal Code that is not enough: no publicly available, signable document for the confidentiality undertaking is evidenced for this route, so the chain stays open. The target list of the specific profile also has to be checked, as it includes non-EU locations via Zurich and London.
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
- Evidenced both region-bound and through the European inference profile - two ways to bind to the EU
- No operator access and no storage in the delivered state
- Multilingual embedding for mixed-language file corpora
Where it stops
- The target list of the eu inference profile includes non-EU states via Zurich and London - the check belongs to the specific profile, not to the model
- A worldwide profile exists for this model as well; without an organisational policy the EU location is a per-call decision
- Embedding is a side role in the search pipeline, not a substitute for a language model
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 |
|---|---|---|
| AWS Bedrock (EU region Frankfurt) | 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 Cohere Embedding carries the sources with their retrieval dates.
Is Cohere 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