Skip to content
JevDirectory.org
CommunityCookbooks & Demos22 starsVerified 2026-09-22

Jev plays StarCraft shareware

A System One harness that plays the original StarCraft shareware mission Strongarm by reading structured game state and issuing mouse and keyboard commands from Jev choices.

Open github.comphyous/tsai-sc
Category
Cookbooks & Demos
Published by
Community
Author
phyous
Added
2026-09-22
Tagscommunitypythongamesagents

Highlights

  • Jev completed Strongarm on development attempt 16, and the captured original victory screen was visually reviewed.
  • The verified run recorded 421 model decisions, 424 API attempts, 9,445,640 input tokens, and a 382.95 ms median latency.
  • The game is paused during state reads and inference, and Jev receives structured observations while screenshots are only for viewers.
  • The model was requested as jev-latest and resolved to jev-1.13.0, with 3,867 game frames captured over 17 min 37.88 sec.
  • Runs the 1998 Windows executable inside BottleShip and needs Python 3.11+, Node/npm, Chrome with WebGPU, and FFmpeg.

Quickstart

bash
python3 -m venv .venv
source .venv/bin/activate
pip install -e .
scripts/setup-runtime.sh
scripts/start-runtime.sh
python -m tsai_sc.run --env-file "$HOME/.config/tsai-sc/env" --run-dir runs/my-attempt --max-requests 2000

Watch out

MIT-licensed harness; BottleShip is Apache-2.0 and StarCraft assets are Blizzard's, not included. Tested on macOS Apple Silicon, and the single successful run does not establish a win rate.

More like this

jev-plays-pokemon-red
6GitHub stars
Pokémon Red under PyBoy where ordinary Python owns the route and arithmetic and Jev is asked only at branches, with each battle turn's faint prediction scored by Brier.
Cookbooks & Demos#community#python#games
Communityjev-plays-pokemon-red
TypeSafe Mario
294GitHub stars
An experimental controller that has Jev choose NES inputs to play Super Mario Bros. from structured emulator telemetry instead of screenshots, logging every decision with probabilities and latency.
Cookbooks & Demos#community#python#games
Communitytypesafe-mario
Laya vs Jev
71GitHub stars
A side-by-side T-Rex game where local Laya runs on Apple Silicon through MLX and hosted Jev uses the TypeSafe API, showing choices, latency, survival streaks, and crash replays on one course.
Cookbooks & Demos#community#python#games
Communitylaya-vs-jev
Back to all resources

From the community

Posts from builders shipping with Jev right now.

Follow @typesafeai

Inferring Jev's internals from 1,000 calls

iwashi / Yoshimasa Iwase
iwashi / Yoshimasa Iwase
@iwashi86
·

Jevの内部アーキテクチャを推測している技術記事(Jev’s Architecture Unmasked)からメモ。 ・本記事はJevのAPIを約1万回の呼び出して、内部構造を推測したもの ・従来の言語モデルを用いた分類やルーティングでは、トークンを1文字ずつ逐次生成するために膨大な無駄な計算コストが発生していた。 Show more

Reply
Read 2 replies

The open System One roundup

小墨同学
小墨同学
@xiaomovps
·

Jev 发布没几天，开源社区已经开始疯狂复刻了🔥 最值得推荐的五个模型： 1、Laya 421M：原生决策模型，支持 Mac 2、Decider-2B：最像 Jev，基于 Qwen3.5 3、NanoJev 0.6B：专门的 Decision Head 4、Reflex：Qwen3.5 + Direct Logits 5、System-One 4B：专门做概率校准 Show more

小墨同学
小墨同学
@xiaomovps

Jev 刚发布没几天，开源社区就出现了同款🔥 Decider-2B模型，是基于 Qwen3.5-2B 做了特殊调整 它和 Jev 模型是一样的 只做选择 评分和判断 不是文本类的 LLM 模型 但两者还是有几个明显区别： 1、模型 Jev：闭源 System One Model Decider：Qwen3.5-2B，约 1.9B 参数，Apache 2.0 开源 2、价格

Image
Reply
Read 6 replies

The launch post

Trading bot, one decision per block

Classifying 1,500 real emails

vogel
vogel
Anomaly
@ryanvogel
·

this model is actually insane at email classification i tested it on 1500 of my own emails to see how well it works and I am blown away

Watch on X
Diogo Almeida
Diogo Almeida
TypeSafe AI
@CompleteSkeptic

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x

Watch on X
Reply
Read 69 replies

Fast browser use with Stagehand

Kyle Jeong
Kyle Jeong
Browserbase
@kylejeong
·

we built blazing fast computer/browser use with Jev + @Stagehanddev. this task cost $0.001 and executed at near instant speed (in a remote browser btw) the loop: observe the page, send a11y tree as state + actions as questions, Jev decides the next action, then Stagehand Show more

Watch on X
Diogo Almeida
Diogo Almeida
TypeSafe AI
@CompleteSkeptic

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x

Watch on X
Reply
Read 30 replies