Vercel's CEO noted that businesses no longer depend on just one AI lab to meet every requirement.

"Last year many teams selected a single lab partner and planned to build all systems on OpenAI or Anthropic," chief executive Guillermo Rauch said during a Monday discussion with TechCrunch.

Yet firms now grasp how each layer of the AI stack functions - from model and harness to data platform, sandbox and gateway - so "every piece is plug and play."

"You can use OpenAI, you can use Anthropic, or you can use Gemini," he continued. Growth stands out with Gemini because those models deliver "awesome price/performance characteristics" at larger scales.

Chinese offerings such as DeepSeek and Z.ai's GLM-5.2 are also gaining quick traction, he noted.

Rauch observed that the prior year focused on "prototyping" AI tools as teams created agents. Today firms face "the realities of agents in production, and some of the challenges."

Vercel, headquartered in San Francisco, operates a cloud platform that lets developers host and release websites and applications.

The executive's remarks arrive while firms realize that AI spending does not always yield added customer value. The era of pushing staff to consume maximum AI tokens has ended, and leaders now seek ways to trim costs or apply AI more effectively.

Coinbase CEO Brian Armstrong posted on X in June that he tested Chinese LLMs such as GLM-5.2 and Kimi AI's K2.7 as standard options because they cost less than offerings from U.S. labs like Anthropic and OpenAI.

He also described model routing, sending engineer prompts to suitable models for each job to avoid costly frontier models on routine work.

Rauch's view on mixing AI lab partnerships mirrors earlier shifts from single cloud vendors like Amazon Web Services or Microsoft Azure toward multi-cloud setups that reduce vendor lock-in and control expenses.