jevland

post author demo

Zero-shot Jev vs an internal fine-tuned classifier

A team pointed Jev at a private benchmark and reports it came within a few recall points of their fine-tuned Qwen at matched precision, while the fine-tune remains faster and has not been replaced.

Private test: about 5 recall points behind a fine-tune, roughly 3x slower, around $70/month at their volume.

author demo — Result shown by its author. About this label

HTTP availability
Not checked
Browser demo scenario
Not checked
Build
Not checked
Live API test
Not checked

What these checks cover →

Report a problem →

← Back to the directory