JevBench, a reproducible benchmark for typed decision models
Tests for AI decision models. Researchers only.
CategoryFor Developers
SourceHacker News (Show HN)
Traction154 pts
First seenSep 22, 2026
Quick facts
- What it is: JevBench, a reproducible benchmark for typed decision models is a for developers tool. Tests for AI decision models. Researchers only.
- Where it was found: Hacker News (Show HN), first seen Sep 22, 2026, with 154 traction when last checked.
- Tags: Open Source.
- Official link: benchmarkheaven.com/jev-models
Featured in
Related tools
🧠 For Developers277 pts
Mini-AGI
An experimental AI model that keeps learning, trained on a normal gaming graphics card.
TechnicalFree & open-source
🧠 For Developers120 pts
TinyAIArena watch AI agents battle it out
Watch AI bots compete against each other in little games.
Some setupFree & open-source
🧠 For Developers53 pts
Durable Actors
Server building blocks for developers. Not really an AI tool for most people.
Some setupFree & open-source
🧠 For Developers56 pts
Doom or Bloom, map your AI worldview
A quick quiz that shows where you land between 'AI will ruin everything' and 'AI will fix everything'.
Beginner-friendlyFree & open-source · Free plan
🧠 For Developers47 pts
jevals
A way for developers to grade AI answers using the Jev decision model.
TechnicalFree & open-source
🧠 For Developers31 pts
Parsimony
Measures how much extra code AI models add when they program. For developers.
TechnicalFree & open-source
Is this your tool? Claim & feature it.