Vercel gateway data shows TypeSafe's Jev reached 35% of paid teams in two weeks

Vercel's October AI Gateway Production Index, covering September, says Jev launched in limited release on September 15 and reached 35% of paid teams within two weeks, nearly double the reach of Sonnet 5.5, the next most popular model. Yet it took only 0.27% of gateway spend, because Jev is a lightweight decision model that picks among a few predetermined outputs, which makes it cheap and fast for classification. Vercel itself replaced GPT-6 Luna with Jev to classify harmful commands. The index also reports that average price per token fell 47% while spend rose about 13 times since December, and that nearly one in three teams changed their most-used lab in September. TechCrunch notes the gateway undercounts labs that sell directly through their own APIs, so the figures are a partial view of the market. For builders, the data suggests small, specialized models are becoming default components in production pipelines, and that model loyalty is weak. Routing layers and evals that make swapping models cheap are worth the investment.