User repository on GitHub
✨✅ The easiest decisions your code will ever make.
- Forks
- 3
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Repos & SDKs
Vibecheck
A Python library that turns Jev-style decision models into ordinary function calls: check returns a bool, classify picks one option, label picks every option that applies, and score rates on a scale, with probabilities behind every answer.
- Stars
- 43
- Language
- Python
- Reading time
- 3 min
- Published by
- Community
- Author
- jlowin
- Last verified
- Oct 2026
Highlights
- Four functions: check answers yes or no, classify picks one option, label picks every option that applies, and score rates on ordered levels.
- Thresholds shape answers: raise the bar with threshold=0.9, or pass a (low, high) pair so check returns None for the unsure middle.
- A batch asks many questions about one piece of data in a single request; 100 questions on a 5,000-token document took about 1% of separate-request tokens.
- The default backend speaks TypeSafe's System One API with TYPESAFE_API_KEY, TYPESAFE_DEFAULT_MODEL jev-latest and a configurable TYPESAFE_BASE_URL.
- Published as vibecheck-py and imported as vibecheck; sync and async variants ship, plus filter and group helpers for lists.
Quickstart
python
from vibecheck.sync import check
review = "vibecheck is awesome!"
if check("Are the vibes good?", review):
print("OK")Watch out
README reviewed 2026-10-17; the excerpt truncates while describing custom backends and lists no version history, benchmark numbers or pricing beyond a fraction of a cent per decision, and the GitHub metadata shows no licence. It wraps decision models including Jev; the digest shows no TypeSafe affiliation.
Questions about this entry
Short answers drawn from the linked project or page.
A Python library from jlowin that drops decision models straight into your code. Each call returns a plain Python value with no prompts to template, no text to parse and no agents to configure; the README calls it the easiest decisions your code will ever make.
The README presents Jev as the example decision model: it answers typed questions in a fraction of a second with a probability attached to every answer. vibecheck's default backend speaks TypeSafe's System One API, uses jev-latest by default, and pins a versioned model like jev-1.13.0 once thresholds are tuned.
Install it with uv add vibecheck-py; the package is published as vibecheck-py and imported as vibecheck. It reads TYPESAFE_API_KEY from the environment, with TYPESAFE_DEFAULT_MODEL defaulting to jev-latest and TYPESAFE_BASE_URL defaulting to https://api.typesafe.ai; keys come from console.typesafe.ai.
The README says any server that speaks TypeSafe's API works with just a new base URL; for example, Vercel AI Gateway serves Jev without a TypeSafe account, using your AI Gateway key and model typesafe-ai/jev. An API key of some kind is still required.
Quick, common-sense judgments about text: intent, tone, topic, urgency, relevance, and whether a statement is true of a document. The README warns that arithmetic, counting and date comparisons belong in code, and that decision models do not generate text, so use a language model to write replies.
Questions are read literally, so edge cases belong in the question or option descriptions; accuracy drops as data fills with irrelevant detail; and when the top two options are nearly tied, the README suggests treating the case as ambiguous and sending it to a person rather than guessing.
Yes. A batch collects several questions about one piece of data into a single request, and a schema (via assess) turns a typed class into a set of questions asked together, filling the instance. The README reports ten questions about one ticket cost about the same as one.
More Jev resources like this
A System One layout optimizer for decision-model workloads: it reorders rows and fields so a prefix-caching backend reuses input computation, then returns per-record decisions in original order. Its bundled backend serves AutoTrust JEV-9B through vLLM.
Jev Repos & SDKs#community#github#python
Community
0814wdwd/solo_jevPython
0814wdwd/solo_jevPythonA Bangla-focused adaptation of Laya, the non-autoregressive typed-decision model: targeted data preparation, fine-tuning and evaluation for classification, routing and scoring on Bangla, Banglish and code-switched text, with the checkpoint still to come.
Jev Repos & SDKs#community#github#python
Community
shamspias/laya-banglaPython
shamspias/laya-banglaPythonHTTP server for the Mica v0.5 4B Preview decision model: GGUF weights run through llama.cpp and answer noul, choice and score questions with calibrated probabilities over the TypeSafe /v1/systemone format.
Jev Repos & SDKs#community#github#python
Community
akivet/MicaPython
akivet/MicaPythonServing code for the Jeff 1.0 family of Jev-style decision models: a state, a typed question and its options go in, and one forward pass returns a probability for each option. Four open models from 4B to 31B parameters are served the same way.
Jev Repos & SDKs#community#github#python
Community
j-geuter/jeff-servePython
j-geuter/jeff-servePythonA frozen decision server that serves Gemma 4 31B — quantized to FP8 at startup — in TypeSafe's /v1/systemone format, reading the next-token distribution instead of generating tokens. It needs one FP8 NVIDIA GPU with 80 GB and has no authentication.
Jev Repos & SDKs#community#github#python
Community
krishna-gogineni-765/deck31bPython
krishna-gogineni-765/deck31bPythonA TypeScript-first toolkit billed as the standard library for System One models: ready-made classify, extract, pickOne, rank and verify methods, a service-agent runtime that runs conversations turn by turn, and a one-message order parser. A Python port is in preview and the project is in alpha.
Jev Repos & SDKs#community#github#typescript
Community
Engid/question-kitTypeScript
Engid/question-kitTypeScriptFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.