Jev probed on 16,379 benchmark requests: not frontier, but bends the price-performance curve
enn_nafnlaus · reddit · 2026-10-04
After two weeks of probing and benchmarking TypeSafe AI's Jev across 16,379 live requests — measuring latency and billing and investigating what it actually is underneath — one redditor's verdict is sober but interesting:
- Marketed as a frontier-class, hallucination-free reasoner built by a ChatGPT co-inventor that's fast and nearly free; in reality it's smaller and humbler, yet genuinely bends the Pareto curve
- Probing suggests it is unlikely to be any preexisting model rebranded
- The report documents weird behaviors users should know, such as the order of choices strongly influencing selection probabilities
Bottom line: not frontier, but useful for a niche nobody else serves quite this way.
More from Models
- Gemini 3.6/3.7 Flash deprecation imminent, Gemini 4 Argon reportedly coming — leslysandra · 2026-10-04
- ChatGPT User Gets Paid Account Banned for 'Recidivism' With No Appeal — anakinimsorry · 2026-10-04
- User finds ChatGPT image generation producing photorealistic nudity for the first time — flowersslop · 2026-10-04
- BOSSFIGHT benchmark: GPT-6.1 Sol scores 67 running a coffee shop, but lays off the harassment complainant — LordKittyPanther · 2026-10-04
- Why chatting with LLMs is exhausting: verbosity, dropped objects, synonym churn — oran_ge · 2026-10-04
- Yacine teases dragging a nonexistent 'Opus 5.5' out of distribution and forcing it to think — yacineMTB · 2026-10-04