After Jev’s release, AWS, Cloudflare and Databricks have each rolled out products or platform features tied to decision models, pointing to faster adoption of the architecture in AI tooling. AWS introduced the open-source Strands Decider 2B, built on Qwen3.5-2B with its text generation module removed and replaced by a dedicated answer-selection component, then fine-tuned with LoRA. In AWS’s tests, median latency for a single decision was about 115 ms on an RTX 3090, while full training took roughly 11 hours. Cloudflare released two open-source models, Clef and Clef-flash, adapted from Qwen3.8-27B and Qwen3.5-9B. In Cloudflare’s tests on Jev’s official benchmark set, the Clef series outperformed Jev in three of four task categories, and Clef-flash posted 38.8 ms median latency versus 524.1 ms for Jev. Databricks, meanwhile, integrated the capability into its platform through ai_decide, which can be called from SQL or a REST API for classification, scoring and agent routing, with direct compatibility for the TypeSafe AI API.
After Jev’s release, AWS, Cloudflare and Databricks have all moved to adopt decision models.
AWS launched the open-source Strands Decider 2B. The model is based on Qwen3.5-2B, removes the text generation module, replaces it with a decision module built specifically to choose answers, and is fine-tuned with LoRA. In AWS’s official tests, median latency for a single decision was about 115 ms on an RTX 3090, and full training took about 11 hours.
Cloudflare released Clef and Clef-flash at the same time. The two models are adapted from Qwen3.8-27B and Qwen3.5-9B, respectively, and both are open source. In Cloudflare’s tests on Jev’s official evaluation set, the Clef series beat Jev in three of four task categories. Clef-flash posted median latency of 38.8 ms, compared with 524.1 ms for Jev.
Databricks took a different route by building the capability directly into its platform. Its newly launched ai_decide can be called through SQL or a REST API for data classification, scoring and agent routing. The interface is also directly compatible with the TypeSafe AI API. Databricks said a single decision can return in under one second.
According to BlockBeats, decision models are quickly shifting from a new concept to a common component in agent infrastructure, as the trend moves from developer-led reproductions to adoption by AWS, Cloudflare and Databricks.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
