CommunityJev Models & Reimplementations
Clef-omni
A multimodal decision model on Workers AI that takes text, images, audio, or video in one request and returns schema-constrained probabilities, built on a 30B mixture-of-experts backbone with about 3B parameters active.
- Published by
- Community
- Last verified
- Oct 2026
- Capabilities
- Open source
Highlights
- Accepts text, images, wav or mp3 audio, and mp4 or webm video in a single request.
- Built on a 30B mixture-of-experts backbone with roughly 3B parameters active.
- Keeps the same typed question and answer shape as Clef and Jev, so only the model field changes.
Watch out
Multimodal input is billed as input tokens, and the model is new, so confirm the current weights, license, and limits on the model page rather than the announcement.
Questions about this entry
Short answers drawn from the linked project or page.
Audio and video input. It accepts text, images, audio in wav or mp3, and video in mp4 or webm in a single request, and returns schema-constrained confidence scores rather than generated text.
Clef-omni runs on a 30B-parameter mixture-of-experts backbone with about 3B parameters active, and it is served through Workers AI with the same typed-question API as the rest of the Clef family.
Through the Workers AI binding or the REST API, selecting the clef-omni model string. It uses the same state and questions shape, so an existing Clef or Jev caller changes only the model field.
Judgments over rich input: classifying or inspecting images, screening clips or recordings, and routing on audio or video evidence, where the answer is a closed choice, a score, or a probability rather than a description.
It is part of Cloudflare's open decision-model line served on Workers AI; check the model page and Hugging Face for the current weights and license, since the family has grown quickly through October 2026.
Images, audio, and video become input tokens, so a request costs more than a text-only one of the same questions. Cloudflare cut Clef-flash input pricing to $0.038 per million tokens in the same window.
No. Jev currently takes text only, so any multimedia has to be transcribed or described first. Clef-omni is one answer to that gap, and its output shape stays compatible with the decision-model pattern.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
LinkedIn: What multimodal decision models actually change, and what they do not.
LinkedInJevDirectory.org
Hacker News: Decision Index 0.2.1 scores open reproductions of Jev on a frozen suite.
jackbravo4 pointsDecision Index 0.2.1 – 70 open reproductions of TypeSafe Jev's Decision ModelRead the thread on Hacker NewsHacker News: Pointer to a DiffusionGemma plus vLLM tutorial for a Jev-shaped classifier.
1 pointHow to build a Jev-style classifier with DiffusionGemma and vLLMRead the thread on Hacker NewsHacker News: A one-token logprob wrapper that also classifies webcam frames.
allanrbo151 pointsA single function Jev-like wrapper for LLMs, including vision modelsRead the thread on Hacker NewsMore Jev resources like this
Cloudflare's first open decision models on Workers AI: Clef, a 27B model that reads text, JSON, images, or video, and Clef-flash, a 9B model for the latency-sensitive hop, each returning a probability per option.
Jev Models & Reimplementations#community#cloudflare#decision-model
Community
Liquid AI's first decision model, a hosted API that adds image input to the Choice, Score, and Noul shape. Liquid reports it matching or beating GPT-6.1 Sol on four of six tested applications at 19x to 200x lower cost.
Jev Models & Reimplementations#community#decision-model#multimodal
Community
17GitHub stars
Jev-compatible decision models with vision input: LoRA fine-tunes of Qwen3.5 at 4B and 9B serving Choice, Score and Noul on the /v1/systemone wire format, with screenshots and photos placed beside text in the state. Runs on your own NVIDIA GPU; weights are non-commercial.
Jev Models & Reimplementations#community#vision#open-weights
Community
Xiaooolong/vevPython
Xiaooolong/vevPython9GitHub stars
A Jev-style typed-decision baseline built on the frozen Qwen3.5-4B checkpoint that accepts a local image path directly, letting the vision-language model score declared options without generating text. Ships scoring modes, a WebGPU browser demo and benchmark results measured on an RTX 3090.
Jev Models & Reimplementations#community#multimodal#open-models
Community
jiangxiluning/Visual-JevPython
jiangxiluning/Visual-JevPythonThe Hugging Face hub for Laya's open System 1 checkpoints: the 421M English model at the repo root, a 322M multilingual build in a subfolder, and the typed-decisions fine-tune, all Apache-2.0.
Jev Models & Reimplementations#community#open-models#local
Community
The vision sibling: the same text-trained System 1 head on Qwen3.8-27B's image encoder, taking images zero-shot, ranked first of 20 vision decision models on the community Jev Decision Index 0.3.
Jev Models & Reimplementations#community#open-models#local
Community
From the Jev guides
Original write-ups on what this entry is for.
Jev takes text only; Clef-omni adds image, audio, and video input and Liquid d1 adds images, each keeping the same typed-answer shape. Where that matters, where it does not, and the costs the vendors underplay.
Jev Models & Reimplementations#multimodal#open-models#decision-model
Read article
One model in September, a dozen by October. Clef, Liquid d1, Perplexity, and OpenAI all ship the same typed-answer shape now. How to read the rival benchmarks and pick on your own task.
Jev Models & Reimplementations#decision-model#benchmarks#pricing
Read article
Hundreds of repositories carry the system-one tag, but the ecosystem is a dozen ideas in different sizes. A tour of the families, the licenses, and what a download does not include.
Jev Models & Reimplementations#open-models#hugging-face#local
Read article
From the Jev community
Posts from builders shipping with Jev right now.