EY создала AI-роутер: система выбирает дешёвую модель под каждую задачу

EY, one of the four largest audit and consulting firms in the world, has built a corporate AI router — an automatic task routing system between language models that selects the cheapest suitable one. The tool was developed in response to the rapid growth of AI spending within the company, Business Insider reported.

Why did EY need an AI router?

EY's AI router addresses one of the key corporate problems of 2025–2026: employees of large organizations default to flagship language models — even when the task does not require it. The result is uncontrolled growth in token spending with minimal quality gain.

EY's system centrally intercepts requests and routes them on the "minimum sufficient model" principle: document summarization is directed to a compact and inexpensive model, legal analysis of a complex contract — to a flagship one. According to industry analysts, the price gap between budget and top-tier LLMs ranges from 10 to 50 times per million tokens. It is precisely this price gap that makes model routing economically justified for organizations with high volumes of AI requests.

How widespread is this practice?

Model routing is not EY's invention: the approach has long existed in open tools like LiteLLM and RouteLLM and is gradually being embedded in enterprise gateways of major AI providers. What is fundamentally new here is that a Big Four firm built its own solution for its own needs, rather than buying a ready-made product from a vendor.

• EY is part of the "Big Four" audit alongside PwC, Deloitte, and KPMG.
• The AI router was developed exclusively for internal corporate use.
• The system automatically selects the cheapest model capable of completing a specific task.
• According to Business Insider, the tool is already in operation within the company.
• The problem of growing AI costs has been recognized as systemic for corporate IT in 2026.

According to Business Insider, EY describes the tool as a response to "growing AI bills" — a phenomenon that has become one of the top items on the CIO agenda.

"This is a problem that is now spreading throughout corporate IT," —

The Next Web provides the context in which EY developed its router.

What this means

EY is demonstrating a new corporate pattern: large professional services firms are no longer simply adopting third-party AI tools and are beginning to build their own AI cost management infrastructure. Model routing is transforming from a niche engineering practice into a basic element of mature enterprise AI architecture. For most organizations, a concrete question arises: buy a ready-made routing solution from a vendor — or build your own, as EY did.