Cloudflare says its new Clef model means humans no longer need to be in the loop for AI agents
Cloudflare's Clef model lets AI agents make structured decisions 10x faster than competitors — without generating text or waiting for human approval.
Why it matters
Cloudflare is shipping a specialized decision model (Clef and Clef-flash) built on Qwen that enables agents to classify and route work in ~39ms without text generation or human-in-the-loop delays. This directly addresses a bottleneck in agentic deployment: latency and cost of traditional LLM reasoning for high-volume routing decisions.
The key facts
6 to knowClef-flash delivers classifications in ~39 milliseconds
10x+ faster than TypeSafe AI's Jev decision model
Built on Qwen, licensed under Apache 2.0
Designed for structured decision-making without text generation
Challenges TypeSafe AI's Jev as a decision-model alternative
Removes human-in-the-loop requirement for agent decisions
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: With Clef and Clef-flash, Cloudflare is challenging TypeSafe AI's Jev decision model. Clef-flash delivers classifications in about 39 milliseconds, making it more than ten times faster than Jev. Both models are built on Qwen, licensed under Apache 2.0, and designed to let AI agents make structured…