Amsterdam-based AI infrastructure company Orq.ai has unveiled a new product called the AI Router. This move highlights a shift in how big organizations manage large language models (LLMs), focusing on better control, cost management, and sovereignty. By making the routing layer a standalone product, Orq.ai aims to give enterprises more flexibility and control over their AI systems.

The Growing Complexity of Managing Multiple Models

As companies move from small pilot projects to full-scale AI deployments, managing multiple models becomes more complicated. New LLMs are released frequently, leading to constant upgrades and testing. Most organizations now run several models at the same time to balance factors like cost, speed, and accuracy. However, this flexibility introduces problems such as fragmented systems, unpredictable costs, and infrastructure decisions that are hard to change later.

Orq.ai believes that the routing layer, which directs requests to different models, has become more than just a technical detail. It’s now a strategic point that can significantly impact performance and costs. The company argues that making the router a standalone product allows teams to better control how requests are handled without being tied to a specific platform or full-stack solution.

Empowering Enterprises with Modular Infrastructure

The decision to separate the AI Router from Orq.ai’s broader platform reflects a demand for more modular AI tools. Instead of adopting an all-in-one system from the start, organizations can now deploy the router independently. This makes it easier to manage traffic, optimize model usage, and adjust infrastructure without overhauling entire systems.

This approach is already being adopted in regulated sectors like finance. For example, Dutch fintech bunq recently replaced its internal routing setup with Orq.ai’s AI Router. According to Benjamin Kleppe, the company’s GenAI Lead, maintaining a custom routing system was becoming costly and resource-intensive. It also lacked sufficient observability and performance insights.

Kleppe explained that working with Orq.ai allowed bunq to switch to a production-ready router that better meets their needs for governance, scalability, and cost tracking. This shift saves resources and improves overall control over AI operations, especially important in highly regulated environments.

Focus on AI Sovereignty and Geopolitical Factors

Beyond cost savings, sovereignty is now a key concern in Europe’s AI landscape. Questions about where inference happens, which providers are used, and how quickly infrastructure can adapt to political or regulatory changes are increasingly important for large organizations.

Orq.ai’s Router enables companies to set routing policies based on factors like location, latency, or cost, without needing to rewrite their applications. This flexibility helps organizations adhere to local regulations and maintain control over their AI infrastructure in a changing geopolitical environment.

In Europe, AI sovereignty is no longer just a policy debate; it’s a practical challenge. Enterprises need tools that let them decide where their AI inference takes place and how quickly they can switch providers or regions if needed.

By offering a modular, policy-driven routing solution, Orq.ai aims to help organizations navigate these complex issues while maintaining agility and control over their AI deployments.

Inspired by

• https://justainews.com/industries/orq-ai-makes-routing-a-first-class-layer-in-enterprise-ai/