Vercel's CEO said choosing one AI lab to partner with is a thing of the past

In a recent interview with TechCrunch, Vercel's CEO Guillermo Rauch emphasized a significant shift in how companies approach their AI collaborations. He noted that businesses are moving away from the previous trend of aligning with a single AI lab to fulfill all their needs. "Last year, many organizations were focused on choosing one lab partner, often selecting either OpenAI or Anthropic as their sole source of AI capabilities," Rauch explained. However, he pointed out that companies are now gaining a deeper understanding of the various components of the AI ecosystem—ranging from models and data platforms to sandboxes and gateways—leading to a more modular approach where each element is interchangeable. He highlighted the growing popularity of Gemini models, citing their impressive price-to-performance ratio, especially as companies scale their operations. Furthermore, he mentioned the rapid adoption of Chinese models such as DeepSeek and GLM-5.2, indicating a broader acceptance of diverse AI solutions. Rauch reflected on the evolution of AI development, stating that the previous year was characterized by experimentation with AI agents. In contrast, businesses are now faced with the practicalities of deploying these agents in real-world scenarios, along with the challenges that accompany this transition. As companies reassess their investments in AI, Rauch noted a shift in mentality regarding spending. The past approach of maximizing AI token usage is giving way to more strategic considerations about cost-effectiveness and efficiency. For instance, Coinbase's CEO Brian Armstrong shared his exploration of utilizing Chinese large language models like GLM-5.2 and Kimi AI's K2.7 for their cost advantages over American counterparts. This trend mirrors the earlier shift in cloud computing, where organizations have increasingly adopted multi-cloud strategies to enhance flexibility and reduce dependency on a single provider, similar to the current evolution in AI partnerships.