Benchmark
Jev benchmark: what the published tests show, plus our own
Every published Jev test we could verify, what each one really measured, and our own reproducible run with the raw data.
BLOG / 5
We run our own tests, publish the raw data, and explain the results in plain language. Start with the Jev benchmark roundup, the Jev vs Laya head-to-head, or OpenAI Decisions API vs Jev.
Benchmark
Every published Jev test we could verify, what each one really measured, and our own reproducible run with the raw data.
Comparison · Benchmark
A head-to-head on the same sampled items, a surprise in Laya's yes/no answers, and a plain guide to which one to pick.
Comparison
OpenAI's Decisions API is now in public beta with a published price and request format. We compare it with Jev feature by feature, work through the cost math, and show how to stay ready for both.
Comparison
Eight decision models launched or grew up in the two weeks after Jev. One table with prices, licences, context, inputs and leaderboard ranks, each labelled with who measured it, and a plain guide to picking one.
Comparison · Benchmark
Our own run of Clef and Clef-flash on Workers AI against Jev on the same items, with every Cloudflare claim labelled as such, and what changes when you switch.