Arize benchmark: 8 decision models tested on hallucination detection and agent routing

Arize AI @arizeaiOriginally posted 2026.10.08

Open original on X

We benchmarked 8 decision models across hallucination detection, 13 @ArizePhoenix evaluator suites, and agent routing. The top decision models are a commodity on accuracy. Jev, Clef 27B and Liquid sit within a point of each other on hallucination, and within the noise of an LLM judge that costs 60 to 400 times more. Full benchmark: https://arize.com/blog/decision-model-benchmark/