Writing C++ with Jev by picking valid next tokens
The author experimented with TypeSafe's Jev decision/classifier model for C++: at each step it was given only grammatically valid next tokens to pick from, avoiding free-form generation.
Jev + Opus 5.5 Coding Pattern Shows Big Gains
A test shows Jev handling step-by-step decisions while frontier models handle planning and coding, cutting time by 51% and cost by 61% while passing all tasks.
Tried the Jev + Opus 5.5 pattern: Jev makes every typed step decision, the frontier model only plans, codes and critiques. Same coding tasks, same harness: Opus 5.5: −51% time, −61% cost Codex: −39% time, −57% tokens Every task still passed. Small sample, strong signal.
· 1 likesOpen on X
More like this
Using Jev for Code Review
The author thinks code review is a really good use case for using Jev.
Fascin: A Jev-Powered PR Triage Tool
A developer shares Fascin, a PR triage tool built with Jev that assesses urgency, change type, and merge readiness, noting Jev handles the actual reasoning.
