Jev directory
CommunityJev Models & Reimplementations
TechnoBaptist/MiniCPM-OCR-Jev
A Hugging Face model repository for TechnoBaptist/MiniCPM-OCR-Jev, a MiniCPM-based OCR model carrying the Jev name. Its published chat template defines image and video vision tokens, XML-style tool calls, and think-tag reasoning output.
- Reading time
- 2 min
- Published by
- Community
- Added
- Oct 2026
Highlights
- Model repository hosted on Hugging Face under the TechnoBaptist organization.
- MiniCPM-OCR-Jev combines a MiniCPM-style chat template with the Jev model name.
- Template renders image items with <|vision_start|><|image_pad|><|vision_end|> vision tokens.
- Template renders video items with <|vision_start|><|video_pad|><|vision_end|> vision tokens.
- System messages containing images or videos raise a template exception.
Questions about this entry
Short answers drawn from the linked project or page.
It is a model repository published on Hugging Face. The listing pairs a MiniCPM-style chat template with the Jev name and an OCR-oriented model title, making it a weights-and-template artifact rather than a hosted service, demo, or end-user application.
The chat template renders image items as <|vision_start|><|image_pad|><|vision_end|> and video items as <|vision_start|><|video_pad|><|vision_end|>, so content lists can mix text, images, and video. The template raises an exception if a system message contains images or videos.
When tools are supplied, the template injects a tools block into the system message and instructs the model to reply with a <tool_call> wrapper holding a <function=name> block and <parameter=name> entries. It also warns that calls must use that exact XML shape.
Assistant messages may carry a separate reasoning_content string. If none is given, the template splits content on the thinking and </think> markers to recover reasoning text, then renders it inside a think block before the final answer.
The tokenizer configuration shown lists a pad token of <|endoftext|> and an unk token of null, alongside an endoftext-style vocabulary and a Jinja chat template rendered for multimodal content.
The provided repository material does not include benchmark scores, calibration figures, latency numbers, or an evaluation table. Anyone needing measured OCR accuracy would have to run their own evaluation against the weights.
Choose it only when you want a MiniCPM-shaped multimodal OCR model under the Jev name and can supply the serving or training stack yourself. Teams needing Jev's typed decisions with calibrated probabilities, hosted inference, or documented benchmarks should look elsewhere.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: Standard One SH (StandardThinking)
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 NewsMore Jev resources like this
Open Jev-style decision model built on LiquidAI's LFM2.5-350M backbone, a clone of kev. Serves the System One API contract so the TypeSafe SDK works against a local server, returning probabilities for typed questions from one forward pass.
Jev Models & Reimplementations#open-models#python#local
Communityfranckverrot/lev
PostHog's reasoning Jev-style classifier with a diffusion drafter, reporting higher accuracy than Kev-9B and hosted Jev on held-out and JevBench tiers.
Jev Models & Reimplementations#community#evaluation#open-models
Community
PostHog/jeevesPython
PostHog/jeevesPythonSystem One models from first principles by Qi Li and Xinchao Wang at the National University of Singapore: give Jeva a state and typed questions and it returns candidate probabilities, a Boolean probability, or an expected rubric score, with reusable option memory.
Jev Models & Reimplementations#community#github#python
Community
liqiiiii/jevaPython
liqiiiii/jevaPythonApache-2.0 Hugging Face repository for RSI-JEV-27b, a 27B multimodal decision model based on Qwen3.8-27B that answers supplied multiple-choice questions over images, returning a selected option and probabilities for all options in one forward pass.
Jev Models & Reimplementations#open-models#benchmark#calibration
Community
Hugging Face model card for JEV-Computer-Use, a System-One decision model for computer and browser use that reads out calibrated discrete action probabilities in a single forward pass instead of verbose text reasoning.
Jev Models & Reimplementations#open-models#browser#agents
Community
Satirical Python script that answers a query with DO or DO_NOT plus a random confidence percentage, presented as a natural alternative to Jev. MIT-licensed, Python 3.8+, one dependency, no API keys or model weights.
Jev Models & Reimplementations#github#python#open-models
CommunityAhmad170412/Koin
From the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.