Jev Cuts AI Decision Costs 100x; Vercel And Cloudflare Rush To Add It

Releases·Tool Release

choosekit runs Jev-style local decisions via llama.cpp: 96.53% accuracy, 239 ms vs Jev's 368 ms

A developer unhappy with sending data to Jev's hosted API built choosekit, a small TypeScript package that extracts choices and probabilities from models running locally in llama.cpp. Tested with Qwen3.8 27B Q4 XL against Jev on SemIf's 144-task benchmark, both hit 96.53% accuracy, with the local setup faster (239 ms vs 368 ms median). The tool targets practitioners who want Jev-style decision extraction without third-party data exposure.

A developer unhappy with sending data to Jev's hosted API built choosekit, a small TypeScript package that extracts choices and probabilities from models running locally in llama.cpp. Tested with Qwen3.8 27B Q4 XL against Jev on SemIf's 144-task benchmark, both hit 96.53% accuracy, with the local setup faster (239 ms vs 368 ms median). The tool targets practitioners who want Jev-style decision extraction without third-party data exposure.

Jev is a demo where an LLM agent plays nine real-time classic games at the same time, driven by a single API call per hour of play costing $1.80. It showcases efficient agentic game-playing — the LLM handles control decisions across parallel game environments. It is of interest to developers exploring cheap, multi-task real-time agents.

Jeda.ai is a platform that uses LLMs to assist in business decision-making, making the process more transparent and inspectable. It helps users understand and verify AI-generated recommendations, addressing the need for accountability in AI-driven decisions.

Databricks announced a 70% reduction in AI coding spend, likely through internal efficiency measures. The news, shared via a Databricks link on Hacker News, highlights cost optimization in AI development.