Can I run it without paying for the Jev API?

Yes. The same call runs on Jev, OpenRouter, Ollaya, zero-shot NLI, Transformers models or any OpenAI-compatible server, with Cascade asking the cheapest first.

How do I avoid acting on borderline judgments?

With ask_human it can return YES, NO or UNSURE, and words like lean and stakes settings adjust how careful the thresholds are.

Will scoring a thousand tickets cost a thousand full calls?

Use gut.each for one decision per subject with concurrent Jev requests, batched local passes and caching of repeated questions.

How do I test code that uses gut without a model key?

Use FakeBackend with fixture answers for tests, and try the browser demo which runs the local model with no key or server.

What other backends can gut run besides Jev?

OpenRouter, Ollaya, a local NLI ZeroShotBackend, a Transformers small LLM, any OpenAI-compatible server such as Ollama or vLLM, and a Cascade that asks the cheapest model first and passes on only what it is unsure of.

What does gut's ask_human option do?

It adds an UNSURE outcome so a borderline answer goes to a person, and you can state care in words with lean="yes" or stakes="high" and gut works out the thresholds.

How do I install gut and what is the package called?

pip install "gutfeel[jev]" for Jev plus OpenRouter or Ollaya, "gutfeel[local]" for the CPU backends, or gutfeel for the core; the PyPI package is gutfeel but you import gut.

CommunityJev Models & Reimplementations

Gut — Kungie

Python helper that turns judgment calls into one-liners like likely, classify and rate, defaulting to Jev while also running on local or OpenAI-compatible backends.

See highlights

Open github.com

Stars: 6
Language: Python
License: APACHE-2.0
Published by: Community
Author: Kungie
Last verified: Oct 2026

Kungie/gut

User repository on GitHub

View on GitHub

Judgment calls as one line of Python, built for TypeSafe AI's Jev and running on any small model: likely / classify / rate → YES, NO or UNSURE. Also local NLI, local LLMs, Ollama, vLLM, OpenAI.

Forks: 0
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.Homepage

This is one of several Jev listings for the same project. The canonical entry is Gut — vinibrsl.

About this project

Judgment calls as one line of Python — built for Jev, and running on any small model. Try it in your browser → The site runs gut's local model in the page: no key, no server. Your code keeps running into questions that aren't logic: Is this comment spam? Which team owns this ticket? How urgent is it? Is the agent's task done? Until now there were three answers: Regex and keyword rules — free and instant, and wrong the moment someone phrases it differently.

Quoted from the project's README. Read the full README.

Tagscommunity github

Works forJev for Classification

Used inJev for Software & DevTools

Languages: Python

Highlights

Exposes likely for yes/no, classify for enums and rate for ordered scales.
Returns YES, NO or UNSURE with word-level lean and stakes threshold control.
Batches with gut.each and semantic calls, with caching and concurrent requests.
Runs on Jev, OpenRouter, Ollaya, zero-shot, Transformers or OpenAI servers.

Watch out

Pre-1.0 helper; Jev is the designed default but other backends have different cost and accuracy trade-offs. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

kev-onnx

Self-hosted CPU-only server that speaks the /v1/systemone wire format for the KEV decision model, packaged as one container with ONNX Runtime and a health check.