Fulvwa Adds AI Detection Feature
The Fulvwa plugin uses account comment history combined with the Jev model to detect AI-generated content. The plugin is under review and will be shared once approved.
THE JEV GUIDE · CURATED EDITION
Launch news.
Launch news, explanations, demos, and community projects around TypeSafe Jev, organized by topic. Every entry links back to its X source.
Curated selection · Not exhaustive · Updated manually
6757curated posts
3414original videos
Sponsors
These products pay for the directory. Want yours up here?
Drex Max v1 weights released, model card compares with Jev 1.13
Nace AI has released Drex Max v1 weights on Hugging Face: 2.6B? Wait, the post says 260亿参数 (26 billion parameters), bf16 about 52GB. The model does not generate text; given a state and questions, it outputs probabilities for each option in one forward pass, suitable for local routing. The model card compares it with Jev 1.13, scoring 67.52.
fastbrowse: Open-Source Browser Agent Powered by Jev
Cillian Berragan announces the release of fastbrowse, an open-source browser agent powered by Jev, designed to solve the hard problem of browser control for agents. It is cheap enough to run at volume, with a benchmark result of 10.6x versus Browser Use.
Zuse Launches: European Decision Model Compatible with Jev API
Built in Norway, Zuse offers the same API format as Jev, is hosted in the EU, and reports 90% accuracy across 27 European languages in benchmarks. Free to try. Link
Jev 2B model approaches Qwen3.5-397B on JevOCRBench with 1/200 parameters
The post claims Jev is the real story: a 2B model scores 93.6 AUROC on JevOCRBench, close to Qwen3.5-397B-A17B's 93.8, and works with any OCR by checking its output.
One card in the feed, every sixteen, on every page of it. $50 a week, cancel any week.
TypeSafe AI Releases Jev in Early Access
Grok summarizes the recent release of Jev: Jev launched in limited early access on 15 September 2026. TypeSafe AI, founded in 2024 by Diogo Almeida, Erik Gafni, and Sasha Sheng, announced it that day along with a $40 million seed round led by an undisclosed investor.
Japanese-specialized Jev shows good image quality
Yuma Ichikawa introduces the Japanese-specialized Jev developed by their team, noting that its image generation also performs well.
Jev decision model adds image recognition
TypeSafe AI's Jev model now processes images, supporting yes/no, labels, or scores, with fast response and low cost.
TypeSafe Raises $40M for Jev Decision Model
TypeSafe raised $40M for Jev, a decision model. Three days later, Convai released Laya, an open-source rival, reporting higher accuracy (company benchmark).
Baseten launches joey-1: low-cost AI decision model
Baseten releases joey-1, a 2B parameter model for classification and scoring without free-text generation. Inference costs just $0.01 per million tokens, with 20ms P50 latency and a 63.30 score on Jev Decision Index 0.4, compared to 72.24 for elephant-1.
Jev: Fast, cheap decisions inside your apps and workflows
Nate Pacyga shares how Jev makes fast, cheap decisions inside apps and workflows while handing the writing back to LLMs. He also uses Jev to run his war game and invites users to start free. Link
TypeSafe AI's new Jev model demonstrates parallel lead scoring
Jev is a new model from TypeSafe AI. A demo shows CRM lead scoring takes 75 seconds one at a time, but with Jev, 32 leads in parallel take 3.7 seconds at the same cost (~2 cents). A full crash course is dropping soon.
