Clef (AI model): news timeline & CVEs

Cloudflare says its new Clef model means humans no longer need to be in the loop for AI agents

Cloudflare launched open-weight Clef and Clef-flash decision models so AI agents can act without a human in the loop.

Cloudflare released Clef and Clef-flash, decision models that return classified answers with probabilities for agent routing rather than free-form text. Clef is based on Qwen3.8-27B and Clef-flash on Qwen3.5-9B; Cloudflare reports median latencies of about 209 ms and 39 ms, versus just over 524 ms for TypeSafe AI's Jev. Both run on Workers AI, are on Hugging Face under Apache-2.0, support a 64,000-token context and images, and can be fine-tuned through a new reinforcement-learning service using Replicate technology.

The DecoderOct 2, 18:19 UTC · 8h agoModel release 5 sources

Cloudflare Releases Clef and Clef-flash: Open-Weight Decision Models That Return Typed Probabilities Instead of Text

Cloudflare released Clef (27B) and Clef-flash (9B), Apache 2.0 decision models that return typed probabilities.

Cloudflare's Workers AI team released Clef and Clef-flash, open-weight decision models under Apache 2.0. Clef is post-trained from Qwen3.8-27B and Clef-flash from Qwen3.5-9B; both return probabilities for typed questions (yes/no, choice, score) rather than generated text. On Cloudflare's Decision Index 0.2.1 shortlist, a Clef model scored highest on 7 of 10 benchmarks, including BANKING77 macro-F1 of 94.20 versus Jev's 79.74, though Jev leads GPQA Diamond, MMLU-Pro, and BBH. Weights are on Hugging Face, both run on Workers AI, and all figures are vendor-reported without independent replication.

MarkTechPostupdated Oct 2, 18:19 UTC · 8h agofirst Oct 2, 01:00 UTC · 1d agoModel release 5 sources

Cloudflare/clef-flash — new model trending #29 on Hugging Face

Cloudflare released Clef-Flash, a 9B multimodal decision model post-trained from Qwen3.5-9B.

Cloudflare published Clef-Flash, a 9B multimodal decision model post-trained from Qwen/Qwen3.5-9B and released under Apache-2.0. It reads a state as text, JSON, images, or video plus a schema of typed questions and returns one probability per allowed option in a single forward pass, with no free-form generation. A joint schema head scores noul, choice, and score questions, and the API matches Jev and SystemOne. On Cloudflare's Decision Index 0.2.1, Clef-Flash scored 98.8 on BFCL and 93.1 on API-Bank, ahead of the larger Clef variant on those two benchmarks.

Hugging Face trending modelsupdated Oct 2, 18:19 UTC · 8h agofirst Sep 30, 21:15 UTC · 2d agoModel release 5 sources

Cloudflare/clef — new model trending #27 on Hugging Face

Cloudflare released Clef, a 27B open multimodal model that answers schema-constrained decisions in one forward pass.

Cloudflare published Clef, a 27B multimodal decision model post-trained from Qwen3.8-27B and released under Apache-2.0. It reads text, JSON, images, or video plus a schema of typed questions and returns one probability per allowed option in a single forward pass, without free-form generation. A joint schema head scores all questions together, and the API is compatible with Jev and SystemOne. On Decision Index 0.2.1, Clef scores 98.5 on BFCL, 69.2 nDCG@10 on ToolRet, and 91.9 accuracy on API-Bank; a smaller Clef-Flash variant is also available.

Hugging Face trending modelsupdated Oct 2, 18:19 UTC · 8h agofirst Sep 30, 21:15 UTC · 2d agoModel release 5 sources

Clef

Timeline4 articles

Appears with

Cloudflare says its new Clef model means humans no longer need to be in the loop for AI agents

Cloudflare Releases Clef and Clef-flash: Open-Weight Decision Models That Return Typed Probabilities Instead of Text

Cloudflare/clef-flash — new model trending #29 on Hugging Face

Cloudflare/clef — new model trending #27 on Hugging Face