Skip to content

Jev directory

Decision Blackjack

A benchmarking studio and real-time simulator that pits Jev (remote TypeSafe System One), a local Laya model and the optimal blackjack Basic Strategy against each other on identical deterministic shoes. It scores every deviation turn by turn and runs headless batches of up to 100,000 rounds.
Stars
3
Language
Python
License
MIT
Reading time
2 min
Published by
Community
Author
mrbanana55
Last verified
Oct 2026

Highlights

  • Three players, one deterministic shoe sequence: every table receives identical generations from a shared shoe pool.
  • Jev gets dynamically generated prompts with the table situation, legal actions and hand values, and returns a structured action via System One.
  • A decision comparison oracle scores every move against Basic Strategy and categorizes deviations by hard totals, soft totals, pairs and surrender.
  • Fast batch mode runs up to 100,000 rounds headless; visualizer mode shows a 2x3 table with dealing animations and a speed slider.
  • Mock mode simulates both AI players when you have no TypeSafe key or local model, and either player can be toggled off.

Quickstart

bash
cp .env.example .env
docker compose up --build

Watch out

README reviewed 2026-10-17; the README shows no results from actual benchmarking runs, so its comparison claims are untested here, and Laya requires a separate local inference server at 127.0.0.1:1234 that the project does not ship.

Questions about this entry

Short answers drawn from the linked project or page.

From 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.

Follow @typesafeai