Jev beats Gemini 2.5 Flash Lite on classifier eval quality and speed
Malte Ubl ran TypeSafe AI's Jev against an existing classifier eval, where it saturated the eval on quality and was 6x faster than Gemini 2.5 Flash Lite.
Entity matching with Jev: a probability-threshold workflow over 33,411 pairs
The author shares a first attempt with Jev: feeding 33,411 pairs to decide whether they describe the same event, using returned probabilities as thresholds—below 0.3 as different, 0.3–0.7 routed to an LLM, above 0.7 as the same—and reports roughly half the time.
איך עשיתי את זה עם Jev? (ניסיון ראשון) נתתי כל פעם זוג ולשאול האם מתארים את אותו הדבר, סה״כ אצלי 33,411 זוגות. הוא מחזיר הסתברות, אז הגדרתי טווח: - אם מתחת ל-0.3 - זה לא אותו אירוע - אם בין 0.3 ל0.7 - לא בטוח, נשלח ל-LLM להחליט - אם מעל 0.7 - זה אותו האירוע ובחצי מהמהירות
· 0 likesOpen on X
More like this
Fast row classification in DuckDB with Jev
The author released a DuckDB extension that uses TypeSafe AI's Jev to quickly classify rows in CSV/Parquet files or DuckDB tables, taking about 10 seconds for 1k rows—faster than an LLM and more ergonomic than a classifier.
Classifying 1,018 AI Papers with Jev for $0.08
The author summarized each paper with DeepSeek V4 Flash, then sent the title, summary, and 24 candidate topics to Jev for classification, processing 1,018 AI papers at $0.08 total cost and 256ms median end-to-end latency per paper.
