AI industry shifts toward multi-model usage and open-source ecosystems

Enterprise artificial intelligence strategies are shifting from a reliance on single frontier models toward multi-model orchestration. Recent data indicates that high-cost frontier models, such as Anthropic’s Fable 5 or OpenAI’s GPT 5, do not always account for the majority of business spending. Instead, organizations are increasingly adopting “model routing,” where tasks are distributed among various models based on complexity, cost, and latency.

Reports suggest that many enterprises now rely on an average of seven different AI models, with over half of organizations using model chaining or orchestration. This approach allows companies to use powerful models for planning and execution strategies while utilizing lower-cost or open-source models for simpler tasks like drafting emails.

In Europe, the focus is shifting toward strengthening the open-source AI ecosystem to maintain technological sovereignty. While closed models currently lead in raw capability due to massive compute and data advantages, open-source and “open-weight” models offer benefits in transparency, customizability, and efficiency. The European Union has integrated open-source AI into its broader digital autonomy and technology strategies to remain competitive.