Updated Nov 10, 2026
855 projects, posts and guides about Jev, the System One model from TypeSafe AI. Each entry links to its source and shows the cost and speed its author reported.
Diogo Almeida
@CompleteSkeptic
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
Newest first, 100–132 of 1072. All 1072 in curated order
Matt Canham
@matthewcanham
I've had many people reach out saying they've read everything they can about Jev but still don't get it. Here's an explanation for the normies
Guidex.com
Matt Canham’s X article, written for everyone who read the launch posts and still did not get it.
GitHubOpen source
Ollama for decision models: pull one by name and serve it locally.
Mert Cobanov
1.3k
Our guidemadewithjev.com
Research, testing, publishing, demand and analysis — each with builds whose authors published the cost. 700 leads scored for $0.09, 100,000 posts analysed for $0.67.
GitHubOpen source
An open decision server with Jev's wire API, on DiffusionGemma 26B.
razorback16
709
GitHubOpen source
Typed probabilities from an open 27B model, with no generated JSON.
Zefan Cai
415
Guidetypesafe.ai
TypeSafe’s launch post: the model, the evals, and their caveats.
GitHubOpen source
Decision models trained by a self-improving loop of AI agents.
Shanghua Gao
117
Our guidemadewithjev.com
Jev in search and RAG, by job: route the query, rerank the shortlist, filter passages before the writer. The per-query costs builders published, and a test it lost.
GitHubSearch
Web search where a decision model picks the sources and ranks them.
superagents-lab
535
GitHubRouting and model choice
One typed Choice question picks the model, tool or subagent.
BillionsBobby
470
Guidegithub.com
The GitHub awesome list for Jev: use cases, projects, SDKs, tools and learning resources, with pricing, limits and a quick start.
GitHubAgents and browsers
Browser use where Jev picks one action per step and code owns the loop.
jkudish
336
GitHubAgents and browsers
Give an agent a model idea and get a LEGO model in LDraw.
anteloc
357Our guidemadewithjev.com
Classify with Jev at the costs builders published: 1,018 papers for $0.08, 1,000 posts for $0.03. Choice, Noul or Score, many labels per call, and when to hand off.
GitHubRouting and model choice
Jev sets the reasoning effort for GPT-6 during Codex tasks.
miuuyy
306
GitHubCoding and code review
Jev ranks each change P0 to P2 and shows you only the P0s.
egma-ai
144
Avid
@Av1dlive
everything you need to start building with jev, in one article. code, architecture, diagrams... everything you need to follow the build and make it your own.
Guidex.com
Avid’s builder’s guide, and the fullest public write-up of the pattern: where the decision layer sits in a coding harness, the five rules that keep it enforceable, and keel, the Rust app it was built in. Honest about the self-improvement part it has not proved.
GitHubCoding and code review
A linter that checks code against plain-language rules.
codegirl-007
126
GitHubRobotics and devices
Ask your house a question, get a number back.
AboveColin
78
Our guidemadewithjev.com
Using Jev in place of an LLM judge: 1,709 editorial judgments for under a cent, a social post judged for $0.00003, and what the confidence score is and is not. Where it beats a generative judge, and where it must not be one.
2BParameters
Strands Agents
9Text and link questions · 8 image questions · 0.45 review threshold
ArticleSecurity and abuse
Every plugin listing is screened before it appears in the catalog.
EmDash CMS
6Models at launch · 5 ms laya, one question · 43 ms openjev 27b, one question
ggml-org
CyrilXBT
@cyrilXBT
Guidex.com
cyril on where the 200× and 400× headline figures come from and why they are a ceiling rather than a typical result: 31 bounded checks counted in one real session, and the decision-point audit that finds yours.
Same directory, different order: the entries added most recently, rather than the curated order the home page opens in. The date on a card is the day it was added here, not the day its author published it — this directory is newer than most of the work in it.