ai
Vercel CEO Says Companies Now Use Multiple AI Models Instead of One Lab Partner
Guillermo Rauch said firms now mix models from multiple providers across the AI stack. The comments reflect efforts to control spending after last year's prototyping phase.
Jul 7, 3:05 AM·1m read2 sources
FlexibilityConsistency
Audio version
Tap play to generate a narrated version.
Developing·Limited corroboration so far. This page will refresh as more sources emerge.
Vercel chief executive Guillermo Rauch said companies have moved away from relying on a single AI laboratory for all development work. In a Monday interview with TechCrunch, Rauch stated that last year many organizations selected one partner such as OpenAI or Anthropic to build everything on.
He said firms now treat each component of the AI stack—model, harness, data platform, sandbox, and gateway—as interchangeable.
"You can use OpenAI, you can use Anthropic, or you can use Gemini," Rauch said. He noted particular growth for Gemini models because of their price and performance when scaled. Rauch added that Chinese models including DeepSeek and Z.ai's GLM-5.2 are also seeing rapid adoption.
He contrasted last year's focus on building AI agents with current work on production deployment and associated challenges. Vercel, based in San Francisco, operates a cloud platform that helps developers host and launch websites and apps. The shift mirrors earlier moves by companies away from dependence on one cloud provider toward multi-cloud arrangements.
Coinbase chief executive Brian Armstrong posted on X in June that his company was testing Chinese large language models such as GLM-5.2 and Kimi AI's K2.7 as lower-cost defaults. Armstrong also described routing prompts to the most suitable model for each task rather than defaulting to frontier systems.
