Cortecs is an AI gateway from Vienna, Austria, that gives developers one OpenAI-compatible API for many language models. Each request is routed to a suitable cloud provider based on cost, speed or a balance of both, and the gateway can fall back to another model if the first one is unavailable. A project setting and a request parameter called eu_native restrict routing to providers established in the EU, and further options control zero data retention and the use of quantized models. Teams can manage API keys, budgets, usage reports and presets in a web console, and bring their own provider keys. Cortecs suits companies that want to use hosted models while keeping control over which providers process their data.
European AI gateways
AI gateways give applications a single API endpoint through which requests are routed to models from many different providers. They manage API keys, usage tracking, budgets and fallbacks, and often add caching, logging and guardrails, so teams can switch or combine models without changing their application code. This category covers European AI gateways that offer an API endpoint hosted in the EU, EEA, EFTA or the United Kingdom and that can be configured to route requests only to European providers whose infrastructure is hosted there. A gateway needs access to several model providers, because routing between them is its core function. It differs from large language model (LLM) APIs, where the provider runs the models itself, while an AI gateway forwards requests to such providers. Plain GPU hosting and services tied to a single model provider are not included.
This category contains services from companies based in a member state of the European Union (EU), the European Economic Area (EEA) and the European Free Trade Association (EFTA).
You can read more about the listing criteria the listing criteria.
European services
Melious is an AI gateway from Saarbrücken, Germany, that gives access to more than 60 open-weight models through one API compatible with the OpenAI and Anthropic SDKs. Requests are distributed across inference providers in eight European countries, and the gateway tries the next provider automatically if one fails. Users choose a routing flavor such as balanced, speed, price, eco or batch by adding a suffix to the model name. Besides text models, the service offers embeddings, image generation, speech models and a chat interface with tools like web search and code execution. Melious does not route to providers outside Europe, so there is no setting to enable for European-only routing.
EUrouter is an AI gateway from Amsterdam, Netherlands, that offers one OpenAI-compatible API for more than 100 language and other AI models. Requests are routed by price, latency and availability, with failover between providers, per-key spend limits and usage monitoring. Its providers include Mistral AI, OVHcloud, Scaleway, IONOS and other European companies, as well as European regions of AWS Bedrock and Microsoft Foundry. Routing options let users allow or exclude individual providers, cap data retention at zero days and use the eu_owned option to limit requests to providers headquartered in the EEA. Developers who already use OpenRouter-style integrations can switch by changing the base URL.
LLMrouter.eu is an AI gateway from Munich, Germany, operated by the IT consultancy MAKONIS GmbH. It offers one OpenAI-compatible endpoint through which organizations reach language models of several providers, with automatic failover between them. Administrators manage users and API keys centrally, and can choose models by provider, region and data center location. Higher plans add a privacy filter and jailbreak protection that check prompts for sensitive content, a manager for EU AI Act documentation and advanced routing rules. The gateway can also be run in an organization's own Kubernetes cluster.