Skip to content
JevDirectory.org
CommunityPractices & Patterns260 starsVerified 2026-09-22

openJev-verdict-2.0

A ModernBERT-based open decision model that evaluates Choice, Score, and Noul schemas in one non-autoregressive pass; Verdict 2.0 reports 77.10% accuracy on 2,000 held-out enterprise decisions.

Open github.comHeman10x-NGU/openJev-verdict-2.0
Category
Practices & Patterns
Published by
Community
Author
Heman10x-NGU
Added
2026-09-22
Tagscommunitypythonopen-modelscalibrationbenchmarks

Highlights

  • Verdict 2.0 uses 149.6M parameters and reports 77.10% accuracy, ahead of Laya's 76.60% and Jev's 72.70% on 2,000 held-out decisions.
  • The 151M Verdict checkpoint is evaluated on 231 public JevBench tasks; inference fixes lifted hard-tier ECE from 0.298 to 0.118.
  • The v1.4 update is inference-only - weights are byte-identical - adding auto-loaded calibrators, NLI-style candidate templating, and a 512-token context budget.
  • Decisions run in roughly 20 to 25 ms and return calibrated confidence; a WebGPU demo tests customer-service, security, and invoice decisions in Chrome or Edge.

Quickstart

python
from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("heman10x/openJev-verdict-2.0")

Watch out

GitHub reports a non-standard license (NOASSERTION) although the README states Apache 2.0. Results are project-local, Verdict 2.0 weights ship as Git LFS pointers, and the project is independent of TypeSafe.

More like this

SemIf
3.6kGitHub stars
An independent baseline that reads typed option probabilities straight from a frozen Qwen3.5-4B's logits in one forward pass, reproducing Jev's interface pattern with open models rather than Jev's undisclosed model or training.
Practices & Patterns#community#python#open-models
CommunitySemIf
Kev
3.2kGitHub stars
Apache-licensed, locally runnable Jev-style decision models (0.8B, 4B, 9B on Qwen3.5) with released weights, training code, a System One-compatible server, frozen eval suites, and a playground.
Practices & Patterns#community#python#open-models
Communitykev
decider
296GitHub stars
An independent open reproduction of the System One model class: Qwen3.5-based 2B and 35B mixture-of-experts models that return typed Choice, Score, and Noul probabilities in one forward pass, with nothing distilled from Jev.
Practices & Patterns#community#python#open-models
Communitydecider
Back to all resources

From the community

Posts from builders shipping with Jev right now.

Follow @typesafeai

A Claude session from 1M to 86K tokens

Alex Volkov
Alex Volkov
CoreWeave
@altryne
·

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

Watch on X
Reply
Read 164 replies

Vercel's fx safety reviewer, 18x faster

Guillermo Rauch
Guillermo Rauch
Vercel
@rauchg
·

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
Read 137 replies

Jev lands on OpenRouter

700 leads scored for $0.09

Beating Gemini Flash Lite on an eval

Browser Use Ultrafast, powered by Jev