JEVThe best of Jev. Real projects, practical guides and ideas from across the internet.
THE OFFICIAL SIGNAL 4 HIGHLIGHTS
Diogo Almeida@CompleteSkepticLAUNCH POSTDiogo Almeida introduces Jev, a structured decision model from TypeSafe. This resource provides an overview of the model's core approach, practical examples, and performance metrics as reported by the author.
TypeSafe AI@typesafeaiOFFICIALTypeSafe officially launches its research lab and introduces Jev to the developer community. This announcement features a launch film by Diogo Almeida and provides foundational context for the project's mission and future direction.
TypeSafe AI@typesafeaiOFFICIALTypeSafe has removed the waitlist for Jev, allowing public access to the platform. Developers can use the official console to integrate and test typed decision-making capabilities within their own software applications.
JEVLAB ARTTypeSafe AI@typesafeaiOFFICIALTypeSafe highlights Jev’s typed-output approach, where applications provide state and questions to receive structured probabilities, choices, or scores. The announcement also details the integration of the Venice API for developers.
JEVLAB ARTStart with the Launch Post. Explore four official signals.
AI-assisted summaries and translations. Check original sources for context and performance claims.
This Jev-powered system automates document processing by analyzing PDFs page by page. It demonstrates a decision-making workflow that identifies which specific pages require OCR, aiming to optimize resource usage during document digitization tasks.
JEVLAB ART
English
OskarThis project introduces a framework for executing end-to-end software tests using autonomous agents powered by Jev. It demonstrates how agentic workflows can be integrated into testing pipelines to automate complex validation tasks.
EnglishTypeSafe Jev as a decision layer for the Pi coding agent: a measured tool-call gate plus jev_ask for typed, calibrated answers.
JEVLAB ARTEnglishPre-commit hook: one Jev call judges whether your commit message matches the staged diff, plus debug leftovers, unmentioned work, and a credential check.
JEVLAB ARTEnglish
trouA tool that processes trading signals to generate intelligent decisions, enhancing system responsiveness.
EnglishThis repository provides a sandbox environment for testing the Jev model. It serves as a practical playground for developers to explore model behaviors and interactions within a structured, type-safe framework designed for experimental AI workflows.
JEVLAB ARTEnglishThis is a LLM Gateway that mimics typesafe ai structured output. Like an imposter Jev.
JEVLAB ARTEnglishThis Neovim plugin utilizes Treesitter to segment code into functions for analysis. It allows users to query their buffer in plain language, with the tool ranking and returning answers via the quickfix list based on its internal scoring mechanism.
JEVLAB ARTEnglishThis tool provides a real-time firewall for AI coding agents by intercepting tool calls. It uses deterministic rules and Jev-based verification to decide whether to allow, block, or request manual approval for agent actions.
JEVLAB ARTEnglishThis project demonstrates a compact Jev-inspired model designed to run locally on a MacBook. It serves as a practical example of lightweight model deployment for developers interested in portable AI execution.
JEVLAB ARTEnglishThis project explores a novel method for conveying human emotions through synthesized sound-based communication. It demonstrates how simple auditory signals can be structured to represent complex feelings, offering a unique approach to non-verbal interaction in digital environments.
JEVLAB ARTEnglishJev Arcade features a 1v1 first-person shooter experience where Jev autonomously manages movement, aiming, and combat actions. This web-based demonstration showcases real-time decision-making capabilities within a competitive gaming environment at a frequency of approximately 9 Hz.
JEVLAB ARTEnglishThis tool provides a management layer for Pi agents to ensure they complete assigned tasks. It monitors active processes and automatically restarts agents that terminate prematurely before finishing their work.
JEVLAB ARTEnglishjevball is a 3D football simulation where each player operates as an independent Jev decision-making agent. The project demonstrates multi-agent coordination in a sports context, allowing users to observe autonomous gameplay or manually control a specific player.
JEVLAB ARTEnglishLocal Telegram channel JSON analyzer for intent, quality, sentiment, and speaker tone.
JEVLAB ARTEnglishThis project demonstrates a method for OMP agents to reduce context usage. The creator claims that utilizing Jev-based compaction can decrease memory requirements by 30-55 percent during operation.
JEVLAB ARTEnglishChrome extension that re-ranks Google results with TypeSafe Jev and folds away sales pages and SEO filler.
JEVLAB ARTEnglishjev-hub serves as a curated directory aggregating long-form articles and demonstration videos regarding Jev, the TypeSafe AI System 1 model. It provides a centralized resource for exploring community-led discussions and visual examples of the framework in action.
JEVLAB ARTEnglish
Ira BodnarThis project utilizes automated data analysis to streamline search engine and geographic optimization tasks. The creator claims this approach reduces operational costs by 90% compared to traditional methods, though these performance figures remain unverified by independent testing.
English
Albiona HotiJev generates 100 images from sources like NASA and The Met based on a single prompt.
EnglishThis project demonstrates how to integrate TypeSafe Jev with Grok Bot to serve as a cost-effective decision-making layer. It provides templates for skill management and usage gating to streamline bot interactions.
JEVLAB ARTEnglishxtags is a browser tool that labels X posts based on their intended call to action. It utilizes Jev, a specialized model that outputs probabilities rather than generating text, to categorize user intent on the platform.
JEVLAB ARTEnglishjselect provides a utility for filtering source-linked evidence while adhering to specific token constraints. It utilizes Jev Noul relevance judgments and local diversity-aware selection methods to prioritize relevant information for AI processing.
JEVLAB ARTEnglishThis project provides a collection of 110 interactive scenarios and games designed for experimentation. It allows developers to test and modify prompts within a structured environment to observe how different inputs affect AI outputs.
JEVLAB ARTEnglishThis project explores agent memory systems by utilizing Jev to curate and retrieve relevant information. It demonstrates a mechanism for managing contextual data, though the effectiveness of the recall process remains subject to the specific implementation details provided by the developer.
JEVLAB ARTEnglishThis project demonstrates evidence-driven frontend quality assurance using Jev Ultrafast and Browser Harness. It provides a synthetic todo application to showcase how these tools can be integrated for automated testing workflows.
JEVLAB ARTEnglishThis unofficial Java SDK provides integration for TypeSafe Jev and Vercel AI Gateway. It demonstrates how to implement these services within Spring Boot applications using WebClient for network communication.
JEVLAB ARTEnglishThis official TypeScript and JavaScript client provides developers with inferred answer types for Jev integrations. It demonstrates how to implement type-safe communication patterns within web applications using a standardized SDK approach.
JEVLAB ARTEnglishTypeSafe AIThis TypeSafe AI reference explains the speculative fan-out pattern. It demonstrates how to improve system responsiveness by initiating multiple concurrent requests and selecting the first successful result to reduce overall latency in AI-driven workflows.
JEVLAB ARTEnglishjev-reviewer is a browser-based tool designed to assist with systematic reviews. It extracts data from trial reports and study documents using standardized templates like RoB 2 or QUADAS-2, providing verbatim quotes and page references for user verification and table export.
JEVLAB ARTEnglishjevprune is a utility designed to filter command-line output for coding agents. It uses specific task descriptions to help agents focus on relevant information by pruning unnecessary data from terminal streams.
JEVLAB ARTEnglishFUn little experiment with Typesafe AI Jev Model playing chess against stockfish :).
JEVLAB ARTEnglishUnofficial Go SDK for the TypeSafe AI System One API — 1:1 parity with the official JS and Python SDKs. Not affiliated with TypeSafe AI.
JEVLAB ARTEnglishThis repository explores running Jev-style parallel typed decisions on consumer-grade 1.5B to 8B parameter models using Apple Silicon hardware. It provides research notes and a demonstration space to evaluate local performance.
JEVLAB ARTEnglishA tool developed using Jev to create a voice builder for Youform, noted for its speed.
JEVLAB ARTEnglishailerix is a type-safe model router designed for Jev systems. It demonstrates a method for banking incoming requests into a structured, typed catalog route to improve system reliability.
JEVLAB ARTEnglishHand the browser work off: an MCP server where a decision model drives the page for your agent, so a flow costs one tool call instead of a turn per click. Ref-based element tables, code-checked assertions, zero-model macro replay. Speaks CDP to your Chrome.
JEVLAB ARTEnglishThis Go client library provides an interface for the TypeSafe AI System One API. It includes optional instrumentation for Langfuse to assist developers in tracking and monitoring their Jev-based application interactions.
JEVLAB ARTEnglishThis project demonstrates an offline decision system utilizing Jev to evaluate various research branches. It aims to streamline complex decision-making processes by providing a structured framework for selecting the most viable paths in research workflows.
JEVLAB ARTEnglishThis project implements a framework for TypeSafe Jev, focusing on System One decision-making and ranking. It provides tools for verification and an opt-in gate for Hermes-based agent interactions within a typed environment.
JEVLAB ARTEnglishAsync Python client for TypeSafe Jev. Typed questions in, probabilities and choices out, no prose to parse.
JEVLAB ARTEnglishThis community-developed Go SDK provides an interface for the TypeSafe System One API. It features typed support for Choice, Score, and Noul questions, while incorporating built-in mechanisms for request retries, context management, and structured error handling.
JEVLAB ARTEnglishTypeSafe AIThis official TypeSafe AI guide demonstrates techniques for extracting structured data from pre-parsed inputs. It provides developers with standardized methods to improve reliability when processing complex information within TypeSafe workflows.
JEVLAB ARTEnglishVoice-driven semantic auto-advance for Slidev, powered by Cloudflare Agents and TypeSafe AI Jev.
JEVLAB ARTEnglishChess where both players are TypeSafe's Jev model: every move is a typed Choice decision.
JEVLAB ARTEnglishJev AI performs quality checks on action labels for physical AI training data at high speed.
JEVLAB ARTEnglish
SaeedA tool built on Jev that identifies programming languages, colors code, and checks lint rules.
EnglishSemIf explores the implementation of semantic conditional logic using open-source models. The project demonstrates how to execute these operations locally on consumer-grade hardware like an NVIDIA 3090 GPU.
JEVLAB ARTEnglishThis experiment compares single-call Jev judgments against multi-dimensional feature extraction across three classification tasks. Results suggest that while decomposition improves performance on Japanese NLI, it also significantly increases the rate of false positives for benign content.
JEVLAB ARTEnglishThis repository provides a community-driven collection of over 110 AI use cases, games, and logic challenges. It features a mobile-friendly interface that allows users to edit prompts and perform A/B comparisons between different model outputs.
JEVLAB ARTEnglishThis tool utilizes Jev to analyze user session replays for identifying software defects. It demonstrates an automated workflow designed to generate bug fix pull requests, aiming to streamline the debugging process for development teams.
JEVLAB ARTEnglishFork of OpenPoke where the yes/no decisions go to Jev instead of Claude Sonnet, with a same-inputs A/B against the replaced LLM decision and a 43,776-request adversarial run on the injection gate.
JEVLAB ARTEnglishJev4Mellea provides a functional adapter designed to integrate Jev with the Mellea framework. This tool facilitates interoperability between the two systems, allowing developers to bridge their respective workflows and data structures within a unified technical environment.
JEVLAB ARTEnglishJev conducts a sentiment-based stock market analysis for the Danish market in 2025, using diverse data sources with low cost.
JEVLAB ARTEnglish
AbolA demonstration where Jev was given $10,000 to trade.
EnglishThis work-in-progress demonstration showcases the Jev model navigating the Minecraft environment. The creator illustrates the agent performing basic survival tasks, such as actively avoiding hostile entities like zombies during nighttime gameplay.
JEVLAB ARTEnglishCommunity Go SDK for TypeSafe AI Jev / System One.
JEVLAB ARTEnglishThis Jev implementation automates support ticket management by classifying incoming requests based on category, urgency, and routing requirements. It demonstrates how to streamline customer service workflows by distinguishing between automated and human-led resolution paths.
JEVLAB ARTEnglishJev-assisted file retrieval and request caching for faster Pi workflows.
JEVLAB ARTEnglish
Aditya SinghThis project demonstrates a voice-controlled interface for Fusion 360. It uses Jev to distinguish between actionable commands and casual conversation, aiming to streamline CAD workflows through natural language input.
EnglishA guide explaining Jev's approach to making typed decisions rather than processing text.
JEVLAB ARTEnglishThis command-line tool integrates with the TypeSafe AI Jev model to perform structured analysis on text. It supports yes/no, multiple-choice, and rubric-based queries, returning calibrated probabilities for use in shell scripts, CI pipelines, and AI agent workflows.
JEVLAB ARTEnglishA stronger one-pass scorer over a variable list of text options: hashed n-gram encoder, rival-aware attention, gated head, temperature scaling, benchmarked against jevlike.
JEVLAB ARTEnglishTypeSafe AIThis official TypeSafe reference document outlines the framework for evaluating model confidence. It explains how developers can measure and interpret uncertainty metrics to improve the reliability and safety of AI-driven decision-making processes in production environments.
JEVLAB ARTEnglishA community post suggesting Jev can be hired as a QA tester through @typesafeai.
JEVLAB ARTEnglishThis project demonstrates a Jev browser agent built in Go that interacts with a live Chrome instance. It provides a framework for automating web-based tasks by bridging the Go runtime with browser automation capabilities.
JEVLAB ARTEnglishThis project demonstrates a Next.js brick breaker game where the paddle movement is controlled in real time by the TypeSafe AI Jev model. The creator reports that the application was developed using Claude Code.
JEVLAB ARTEnglishThis resource demonstrates how to access Jev via the Vercel AI Gateway. It outlines the integration process and discusses potential business applications for the platform, providing a practical guide for users looking to implement Jev in their workflows.
EnglishJob Risk Analyzer provides a CLI and REST API interface to evaluate professional roles. It uses Jev to score occupations based on their potential exposure to AI-driven workforce displacement and their relative resilience in an evolving labor market.
JEVLAB ARTEnglishThis resource showcases an implementation of computer use built on Jev. The creator claims significant performance improvements over Opus 5, noting increased speed and cost-efficiency, while highlighting the system's ability to generalize across different operating systems.
JEVLAB ARTEnglish
Jozef GhermanA free AI tool built on Jev that quickly scans text for low-quality content, processing around 10,000 words in 2 seconds.
EnglishComment-moderation playground: paste a comment, Jev decides what to do with it.
JEVLAB ARTEnglish
Submit a find
한국어
日本語
Español
Français
Deutsch
简体中文
繁體中文
Português
Bahasa Indonesia
ไทย
Italiano
Tiếng Việt
Türkçe
Polski
Русский
4 HIGHLIGHTS
Decisions & routing
Agents & browsers
131
Tools & apps
Research & data
Creative & growth
Games & real time
Start here
Markets & experiments
