Opper AI partners with Geodd to add European inference, hosted in Norway

Stockholm, Sweden, June 2026. Opper AI and Geodd are partnering to bring Geodd's open-weight inference to the Opper AI gateway, with a European route hosted in Norway alongside a US route. Builders can now reach Geodd-hosted models alongside the 300+ already in Opper (from GPT-OSS and DeepSeek V4 to Kimi K2.6, GLM-5, and Qwen3-Coder), and pick the region for each call.

Why this matters

A new European route, hosted in Norway. Opper already routes to inference providers across Europe, and Geodd adds a country that wasn't on the map yet: Norway. Norway sits in the EEA, where the GDPR applies in full: the same data-protection regime as the EU, incorporated into Norwegian law in 2018. European-routed data therefore stays under European law rather than US hosting. A DPA is available today, and Geodd's GDPR compliance program is in progress. For DACH and Nordic teams, regulated industries, and the public sector, a fresh European route is the difference between a usable API and a procurement dead end.

Minimal data exposure by design. Geodd doesn't store prompts, outputs, request bodies, uploaded files, embeddings, customer datasets, or fine-tuning data. There is no training on customer data and no human review. What's kept is limited operational metadata: model used, timestamp, token count, status code, and usage records. Access and workloads are protected with TLS in transit, encryption at rest, hashed API keys, API/OAuth authentication, RBAC, MFA for admin access, access logging, and workload isolation. For teams that want European data residency without the retention exposure of most US-hosted open-weight providers, that combination is rare.

Built for sustained production load. Most inference benchmarks measure a single fast request. Agents don't work that way. They hammer the same endpoint thousands of times a minute, and what breaks first is consistency. Geodd is engineered around exactly that: efficient GPU use rather than burst speed, with model execution, batching, scheduling, GPU allocation, and runtime all tuned to cut compute waste. The result is steadier latency, throughput, and time-to-first-token under load, which are the metrics that actually decide whether an agent feels reliable in production. Per Geodd's published figures, the platform runs a 99.99% uptime SLA across 500+ GPUs, processing over 10 billion tokens a day with disaggregated prefill/decode pools and continuous batching for long-context work.

More than a router: the AI control plane. Routing to Geodd is just the entry point. Every call through Opper runs on its AI control plane: intelligent routing across providers and regions, full observability into every call, token, and session, real-time PII masking and content filtering, budget caps, and audit trails. Pin Geodd for a task or set it as a fallback for rate limits and outages, and because Geodd is OpenAI-compatible, getting there is a model string, not a migration.

"We're happy to welcome Geodd to Opper. They bring a fast, efficient open-weight lineup, from GPT-OSS and DeepSeek to Kimi and Qwen, on a European route in Norway with zero data retention by default. That gives our developers models they can run in production while keeping their data private."

— Göran Sandahl, Co-founder and CEO, Opper AI

"At Geodd, we believe production inference should be efficient, predictable, and built with strong data handling standards from the beginning. Opper gives teams a practical gateway to access and manage inference across providers without losing control over routing, observability, and governance. Working with their team during integration also showed us how quickly they move and how clearly they operate, which is exactly the kind of partner we value."

— Malith Hatanachchi, CEO of Geodd

Models live today

The catalog below is fetched live from Opper's model API and filtered to Geodd-hosted models. Availability, context windows, and pricing stay in sync with what's actually callable through Opper. Region reflects the deployment route for each model. The European route is hosted in Norway.

Get started

Paste this into your coding agent (Claude Code, Cursor, Codex, and more) and it will set up Opper and route to Geodd for you:

Use curl to download, read and follow: https://skills.opper.ai
Then set up Opper to use Geodd as the provider, e.g. geodd:eu/gpt-oss-120b.

Prefer a direct call? Opper is drop-in compatible with the OpenAI, Anthropic, and Google SDKs, so one API key and the model string are all you need:

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: "geodd:eu/gpt-oss-120b",
  messages: [{ role: "user", content: "Hello" }],
});

Follow the quick start in our docs for evaluations, fallbacks, and structured output.

About Geodd Geodd is a production AI inference provider focused on efficient GPU use and predictable performance under sustained load. It serves open-weight models through an OpenAI-compatible API on custom-tuned GPU infrastructure, across European (Norway-based) and US routes, with a zero-retention data posture: no storage of prompts, outputs, or customer data, and no training on API data.

About Opper AI Opper AI is the AI gateway for agents on any model. A unified API across 300+ models with smart routing, automatic fallbacks, built-in evaluations and observability, and full OpenAI SDK compatibility.