Cloudflare has announced two new models named Clef and Clef-flash, according to a report by The Decoder. The release is positioned as a direct challenge to TypeSafe AI and its Jev decision model. Built on Qwen and licensed under Apache 2.0, Clef and Clef-flash are engineered to allow AI agents to make structured decisions without generating text, achieving speeds of about 39 milliseconds for the flash variant.
The Shift Toward Automated Decision-Making
The core proposition behind these new models is the elimination of humans from the loop for AI agent workflows. By focusing entirely on structured decision-making rather than standard text generation, infrastructure providers aim to reduce latency and overhead in automated systems. The announcement noted that Clef-flash operates at more than ten times the speed of competing models like Jev, potentially changing how real-time agentic tasks are handled at the network edge.
What Remains Unclear
Despite the initial details provided in the announcement, several key questions remain unanswered. The disclosure did not specify production deployment timelines, pricing structures, or exact accuracy benchmarks compared to existing decision models. Additionally, the broader integration path within Cloudflare’s existing developer ecosystem was not detailed, leaving developers waiting to see how these models will be accessed in practice.
Further technical specifications and real-world performance metrics will determine whether Clef and Clef-flash can successfully displace current alternatives in production environments.
Lead image: Christina Morillo / Pexels
