EU data residency

Llama models hosted in the EU

Where each Llama model runs in the EU, what the provider keeps and what it costs, route by route.

Get your API keyRun Llama models on Opper
3 of 13 models5 routes5 providers3 countries

Every EU route for Llama

One row per model and route, with each model's best route first. Token prices are per million tokens.

ModelProviderRegionZero data retentionTrainingInputOutputCache read
Llama 3.3 70B
Meta
Regolo
ItalyNo$0.68$3.07Input price
Llama 3.3 70B
Meta
GreenPT
EUNo$1.25$1.25Input price
Llama 3.3 70B
Meta
Scaleway
FranceNo$1.02$1.02Input price
Llama 3.3 70B Instruct
Meta
evroc
SwedenNo$1.14$1.14Input price
Llama 3.1 8B Instruct
Meta
Greenference
EU multi-regionNo$0.01$0.02$0.0050

Llama models not in the EU yet

These Llama models run on Opper outside the EU for now, in the region shown. This page updates within the hour when one gains an EU route.

Llama 3.1 70B InstructUSLlama 3.1 70B Instruct TurboUSLlama 3.1 8B InstantUSLlama 3.1 8B Instruct TurboUSLlama 3.2 11B Vision InstructUSLlama 3.3 70B Instruct TurboUSLlama 4 Maverick 17B 128E InstructAustraliaLlama 4 Maverick 17B 128E Instruct FP8USLlama 4 Scout 17B 16E InstructUSLlama Guard 4 12BUS

Call Llama only in the EU

Every route has its own model id. Call that id on any plan and the request is served on that route, in that location. The examples below call Llama 3.3 70B on Regolo in Italy.

Set it up with your agent

Copy this into a coding agent like Claude Code, Cursor or Codex and it will wire up Opper and call this route.

Or call it directly

import OpenAI from "openai";
const client = new OpenAI({
  apiKey: process.env.OPPER_API_KEY,
  baseURL: "https://api.opper.ai/v3/compat",
});
const completion = await client.chat.completions.create({
  model: "regolo/llama-3.3-70b-instruct",
  messages: [{ role: "user", content: "Hello" }],
});
console.log(completion.choices[0].message.content);

Lock it in with a rule

A model access rule with Inference location: EU, set for your organization or a project, means no LLM or embedding call can reach a route outside the EU, even by mistake. Rules are part of the Control Plane. How model access rules work

Who you contract with

Company
Opper Technology AB, Sweden, reg. no. 559446-3720
Certification
ISO/IEC 27001:2022
Processor
Opper, for every modelSub-processor list
DPA
One DPA covers the whole catalogueRead the DPA

Other model families in the EU

Claude11 modelsOpenAI25 modelsGemini11 modelsMistral28 modelsDeepSeek7 modelsQwen19 modelsGLM10 modelsKimi4 modelsGrok4 models

Frequently asked questions

Is Llama available in the EU?

+
Yes. 3 of the 13 Llama models on Opper run on EU routes, on Regolo, GreenPT, Scaleway, evroc and Greenference in Italy, France, Sweden and the EU multi-region. Call the route's model id and the request is served in the EU.

Which Llama models run in the EU?

+
Llama 3.3 70B, Llama 3.3 70B Instruct and Llama 3.1 8B Instruct. Not in the EU yet: Llama 3.1 70B Instruct, Llama 3.1 70B Instruct Turbo, Llama 3.1 8B Instant, Llama 3.1 8B Instruct Turbo, Llama 3.2 11B Vision Instruct, Llama 3.3 70B Instruct Turbo, Llama 4 Maverick 17B 128E Instruct, Llama 4 Maverick 17B 128E Instruct FP8, Llama 4 Scout 17B 16E Instruct and Llama Guard 4 12B.

Is Llama GDPR (DSGVO) compliant in the EU?

+
Running Llama on an EU route keeps processing in the EU, which settles where your data goes. The rest depends on your own processing and the service's terms. On Opper, one Data Processing Agreement covers Llama and every other model, Opper is your single sub-processor, and each route lists what the provider keeps and whether it trains on your data. Opper is certified against ISO/IEC 27001:2022; the controls, policies and sub-processors are on https://trust.opper.ai. Read the DPA

Does Llama cost more in the EU?

+
Sometimes. Comparing input prices, of the 2 Llama models that also run elsewhere, 1 costs less and 1 costs more on their cheapest EU route. Other hosts in Europe set their own prices, so for open-weight models the difference depends on the host. Every route's own price is in the table on this page.

How do I make sure Llama calls never leave the EU?

+
On any plan, call an EU route's model id and the request is served there. To lock it in, set a model access rule with Inference location: EU for your organization or a project, and no LLM or embedding call to Llama can reach a route outside the EU, even by mistake. Rules are part of the Control Plane. Model access rules

Read this list from an agent

This list as data. The same routes as JSON at opper.ai/models/eu.json (model id, provider, country, zero data retention, training and price per route) and as markdown at opper.ai/models/eu/llama.md.

The whole catalogue. The model catalogue is public JSON at https://api.opper.ai/v3/models, no API key needed. Filter it with query parameters, for example ?type=llm&limit=3, and page through it with limit and offset. Use it when you can fetch a URL but cannot connect to an MCP server. Coding agents that use MCP can connect to https://api.opper.ai/mcp. Two of the server's tools need no account and no sign-in: list_models searches the model catalogue by name, type, provider or capability, and get_guide returns short setup guides. Everything that touches your account needs you to approve the connection in the browser first. About the MCP server

Run your AI in the EU

One API key for every EU route and one DPA for every model, with each call served where you choose.

Get startedRead the DPA