How Jev works: teardown of TypeSafe's typed decision model, replicated on open Qwen models
vykthur · x · 2026-09-25
Victor Dibia published a deep teardown of TypeSafe AI's Jev, a 'System One' decision model that outputs typed, probabilistic decisions instead of text, claiming 70-500ms latency and so-cheap inference that output tokens aren't metered.
- How it works: Jev turns a decision task into a completion prompt — a state plus a list of options, with the prompt ending exactly where the answer goes — returning typed decisions with probabilities rather than free-form text.
- Hands-on experiments: The author replicated the approach on open Qwen models to test whether completion-style prompting works for classification-style decisions.
- Tradeoffs covered: architecture differences vs. autoregressive models, when to use it, and what you give up (generative capability) for low latency and cost.
He also experiments with an interactive-explainer format for his articles, hosted on his own site.
More from coding & agent
- The decade of agents: simulate real customers to test your agent before launch — EXM7777 · 2026-11-11
- FluentCRM MCP Server lets you manage marketing automation from Cursor or Claude — modelcontextprotocol · 2026-11-11
- Databricks' revised Genie agent tells CEO Ali Ghodsi the unvarnished truth — TechCrunch AI · 2026-11-11
- Elvis Saravia: agents will subsume traditional software — and can even cut code bloat — omarsar0 · 2026-11-11
- Every launches Everycut, an agent-native Mac video editor where every control is an agent tool — every · 2026-11-11
- Anthropic's Fable 5.2 built a professional video editor overnight, unsupervised — every · 2026-11-11