AI Weekly Malaysia

AWS released Strands Decider 2B, an open-source 'decision model' inspired by TypeSafe's Jev, the same week OpenAI announced a comparable offering. Built on the torso of Qen3.5-2B, it doesn't generate text — it picks between pre-decided options and returns a calibrated confidence score, and it's small enough to run locally. Amazon distinguished engineer Marc Brooker built an early version after seeing Jev; it briefly topped the Jevbench ranking for models of its size before AWS cleaned it up and shipped it via Strands Labs.

Why: If your agent workflow uses a full LLM call just to answer 'what do I do next?', a 2B local decider with confidence scores can replace that step with lower latency and no per-call API bill — and because the answer domain is closed, you can gate actions on the confidence value instead of parsing free text. Worth benchmarking on your own routing steps before assuming it beats your current prompt.

Cloudflare released two Cloudflare-trained "decision models" — Clef and Clef-flash — hosted on Workers AI, open-sourced on Hugging Face under Apache 2.0, and made Jev-API compatible with Typesafe AI's Jev System One. Decision models return bounded, typed outputs with probabilities (e.g. 95% fashion, 85% ecommerce, <1% phishing) instead of open-ended text, and Cloudflare says Clef currently leads the Jev Decision Index. Cloudflare also debuted an RL product for fine-tuning Clef, and reported its own Threat Intelligence workflow classified a domain in 2.2s with Clef versus 4.7s for gpt-oss-120b, which returned only two classifications.

Why: If you are routing tickets, escalations, or domain/page categories inside an agent loop, a classifier that returns typed labels plus probabilities lets your code branch deterministically instead of parsing LLM prose — and since Clef is Apache 2.0 on Hugging Face you can self-host and test it without committing to Workers AI billing. Treat the 2.2s vs 4.7s figure as vendor-reported on Cloudflare's own Threat Intelligence workflow, so benchmark it on your own inputs before swapping out a prompt-based classifier. The new RL fine-tuning option is the piece to evaluate if your label set is domain-specific and you don't want to retrain a full classifier each time categories change.