Vals AI: Jev matched GPT-6 Astra at 1/500th the cost
An independent test put Jev at GPT-6 Astra's 97.5% accuracy on 400 claim-verification questions for about 1/500th the price. That is the first third-party number big enough to decide a build-or-buy on.
Someone measured it
Vals AI says Jev matched GPT-6 Astra accuracy at 1/500th the cost
Vals AI says it independently tested Jev against 11 other models on 400 claim-verification questions, and it matched GPT-6 Astra's 97.5% accuracy at roughly 1/500th the cost. Separately, OpenRouter launched decision model rankings showing spend and token share by task type, and says TypeSafe is leading all categories today.
On 400 claim-verification questions, it matched GPT-6 Astra’s 97.5% accuracy at ~1/500th the cost.
TypeSafe answers the Jev-killer talk with Nilforoshan's numbers
TypeSafe AI reposted Hamed Nilforoshan's Decisions API benchmark with the line that rumors of its death were exaggerated. Nilforoshan ran the two on predicting how relevant a user query or resume is to a job description, on a service he says serves 2.5 million users, and reported OpenAI 2x more expensive and 5 to 10% worse.
The rumors of our death have been greatly exaggerated
Ramp data shows Jev gaining a point of market adoption in a month
Ara Kharazian posted Ramp's October top SaaS vendors and says Jev captured an entire percentage point of market adoption within a month, which he calls unheard of growth. He frames it as part of cheaper models and routers pulling AI spend down, against image and video models pushing it up. TypeSafe quoted it with "fastest growing".
Jev captured an entire percentage point of market adoption within a month (unheard of growth)