Open-Source Models Now Handle 62% of Vercel Token Consumption
Open-source AI is taking market share faster than expected. The share of open-source models in token consumption on Vercel jumped from 28% to 62% in just two months, according to investor Gavin Baker.
The reason is increasingly clear: closed models remain strong but expensive, while open ones are rapidly closing the quality gap. Open source adds another advantage — control: models can run locally, keeping data in-house with no dependence on third-party APIs, pricing, or access policies.
The price/quality/control ratio keeps improving. If the trend continues, the question shifts from “why run AI locally?” to “why hand over data and pay for a closed API when open source is already good enough?”
Related: Nvidia Puts $7 Billion Behind Open-Weight Models to Rival DeepSeek and Kimi, MiniMax M3: Open-Weight Model with 1M Context Outperforms GPT-5.5 and Opus on Agentic Benchmarks