Vercel data published on Thursday shows that Jev, a decision model from TypeSafe AI, has become the fastest-adopted AI model on Vercel‘s platform by some measures. Within two weeks of its launch, 35% of paid teams were using Jev, nearly double the reach of the second most popular model, Sonnet 5.5. According to Vercel, that is the fastest adoption any model has seen in its first two weeks, making the release a major accomplishment for TypeSafe AI.

Jev was released in limited form on September 15. It is explicitly designed as a lightweight tool for choosing between a few predetermined outputs. By limiting the output space, the model can run cheaper and faster than alternatives like Sonnet, which makes it well suited for classifying large amounts of data quickly. The approach has generated considerable buzz since the release, inspiring competing offerings from OpenAI and Amazon among others. But until now there has been little hard data to back up the enthusiasm.

The numbers come from Vercel‘s AI gateway, a unified API service similar to OpenRouter. The dataset is a necessarily limited view of the AI market, particularly at risk of undercounting major labs such as OpenAI and Anthropic that operate their own APIs. Still, the report shows the buzz around Jev translated into actual usage on Vercel‘s platform. Vercel itself was an early adopter, using Jev to replace GPT-6 Luna for classifying harmful commands.

The same report also documents continuing momentum for open source models, which now account for more than 70% of the tokens processed through Vercel‘s gateway and nearly 25% of total spend. While open source is capturing a larger share, the overall market is growing quickly: spending on frontier labs has grown tenfold, even though it represents a smaller share of overall spend. That combination lends weight to the idea that frontier and open source models may turn out to be complementary rather than competitive.