Sovereign AI. One API.

Run open models on European infrastructure through an OpenAI-compatible API. You choose where your requests may run, and every response tells you where it did.

How it works

Point your existing OpenAI client at the Eldrim endpoint and use your Eldrim key.
Your account carries a residency policy: Sweden only, or EU.
We send each request to an approved European provider that meets the policy. If that provider fails, the next one takes the request.
Each response names the provider and the jurisdiction it ran in, and we keep the same record for your audit trail.

curl https://api.eldrim.net/v1/chat/completions \
  -H "Authorization: Bearer $ELDRIM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "llama-3.3-70b",
       "messages": [{"role": "user", "content": "Hej"}]}'

# Response headers include:
# x-eldrim-provider: ...
# x-eldrim-residency: se

Ways to buy

Option	For
Pay as you go	Teams that want to start today. Priced per million tokens, per model.
Reserved capacity	Production workloads that need predictable throughput and price over months or years.
Private Swedish capacity	Organisations that want dedicated hardware at a named Swedish site. We are looking for design partners.

Where the compute comes from

Eldrim is in early access. Today, capacity comes from established European providers. Our own site in Haparanda, in northern Sweden, is in a feasibility study. When it opens, it becomes one more route behind the same API, and your code does not change.

Early access

Tell us which models you run, how much, and where they must run. We reply with a price.

Contact: to be announced.