Skip to content

jevos

A local yes/no decision model that speaks Jev's wire format on a laptop CPU, refusing Choice and Score until those question types ship.

Category
Models & Reimplementations
Format
—
Published by
Community
Author
feder-cr
Added
2026-09-28
Last verified
2026-09-28

Highlights

  • Yes/no only: choice and score questions are refused with HTTP 422.
  • On 2,000 unseen yes/no items the README reports accuracy 0.815, against 0.927 for Jev and 0.489 for Laya.
  • Short requests are listed at 54 ms on CPU; three questions sharing one state take about 165 ms.
  • Context is 8,192 tokens. Optional noul criteria are accepted and not read.
  • A bundled /dino page plays a Chrome Dino-style game from two yes/no questions per step.

Quickstart

bash
uv run jev serve --gguf jevos-q4_k_m.gguf --device cpu --threads 16
curl http://127.0.0.1:8017/v1/systemone -H 'Content-Type: application/json' -d '{
  "model": "jev-latest",
  "state": "I was charged twice for the same order.",
  "questions": {"billing": {"type": "noul", "instructions": "Is this a billing problem?"}}
}'

Watch out

MIT-licensed. Independent of TypeSafe. Multiple choice and scores are marked soon, and the published accuracy is the project's own comparison.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

X: deepseek-v4.1-flash-jev

Related terms

Glossary definitions related to this entry.

More like this

689GitHub stars
A local server that reads next-token probabilities from a fine-tuned Spark model and returns Choice, Score, and Noul answers with zero generated tokens.
Models & Reimplementations#community#python#open-models
39GitHub stars
A local /v1/systemone server that reads label probabilities from a GGUF chat model, with prompt repetition and a confidence shrink toward uniform.
Models & Reimplementations#community#python#open-models
7GitHub stars
A Python server in front of llama-server that reads one-token label probabilities and exposes them as POST /v1/systemone.
Models & Reimplementations#community#python#open-models

From the guides

Original write-ups that draw on this entry.

A base URL is the whole client change. What still works when Ollaya, Lichen, verdict, Rizzo Flow, or jevos answers POST /v1/systemone on your machine.
Models & Reimplementations#open-models#local#api
Read article
Back to all resources

From the community

Posts from builders shipping with Jev right now.

Follow @typesafeai

Fast browser use with Stagehand

we built blazing fast computer/browser use with Jev + @Stagehanddev. this task cost $0.001 and executed at near instant speed (in a remote browser btw) the loop: observe the page, send a11y tree as state + actions as questions, Jev decides the next action, then Stagehand Show more

Diogo Almeida
Diogo Almeida
TypeSafe AI
@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

Reply

LLM-as-a-judge, sped up

Jev has spoken. It picked which model is AGI. 20–200x faster. 40–400x cheaper. This could make things like LLM-as-a-judge insanely fast and nearly free. (I tried a bunch of prompts and still didn’t burn through $0.10.)

Image
Diogo Almeida
Diogo Almeida
TypeSafe AI
@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

Reply

Instant compaction with Jev

A Claude session from 1M to 86K tokens

This is actually insane. This uses @typesafeai Jev model, as a plugin in Claude to review all the un-nesseasary tool calls, and it takes 1s to run! Like, literally, 1 second to take my Claude session from nearly 1M to ... 86K tokens! 😮 Ask your claude to install it and be  Show more

Image
Image
tamara
tamara
@tamarajtran

found the perfect use case for @typesafeai Jev: instant compaction in 2026, why is compaction still a summarization prompt? Jev can make it instant by scoring every tool call and dropping what’s irrelevant

Reply

Vercel's fx safety reviewer, 18x faster

We're seeing extraordinary results from @typesafeai. Default mode in 𝚏𝚡 is auto, with a safety reviewer analyzing every command. That reviewer runs on GPT Luna today. Jev is up to 18x faster (p95) *and* more accurate. It's coming to @vercel AI Gateway and likely new default.

Pranit
Pranit
Vercel
@fazxes

We benchmarked fx auto mode (safety) classifier with @typesafeai's Jev. tl;dr: ~5-18x faster and more accurate than 𝚐𝚙𝚝-𝟻.𝟼-𝚕𝚞𝚗𝚊, our current top choice

Image
Reply

Jev lands on OpenRouter