jev on Hacker News

97 titles from 80 authors in the last 30 days, with 89 distinct links and 23 Show HN.

Counts titles, it does not confirm facts. Open the source before concluding anything.

Projects on this term

The score comes from our model.

• Show HN: Gutsy – Tiny model for typed decisions on CPU (Jev style) · score 7.5

  Gutsy is a tiny 0.8B-parameter language model fine-tuned for typed decisions (yes/no, choice, score) that runs locally on CPU via llama.cpp. It outputs calibrated probabilities, follows the Jev-style API, and competes with larger models on decision benchmarks while being private and deterministic.

• Show HN: SuperLocalMemory" AI brain that never bluffs: Laya, Jev, Google Gemma · score 7.3

  SuperLocalMemory is a local-first, governed memory layer for AI agents that stores decisions and documents, retrieves them with confidence tiers (including an honest 'not found' response), and works across multiple agents via MCP. It combines on-device embeddings, optional local vision models, and a mesh for sharing memories, aiming to prevent agents from hallucinating memories.

• Show HN: THX-01 – Open- source decision model that runs on phones and beats Jev · score 7.3

  THX-01 is a 322M parameter non-autoregressive decision model that answers typed questions (choice, number, excerpt, citation) over documents in a single forward pass with calibrated probabilities. It uses reinforcement learning with a strictly proper scoring rule to ensure honest confidence, enabling selective automation. Benchmarks show it matches or beats larger LLMs on classification and extraction tasks while running in ~10ms on one GPU.

• Show HN: I made a desktop pet that uses LLM for voice and Jev for poses · score 6.8

  PetPaw is a macOS desktop pet app that uses LLMs for voice interaction and pose selection, with support for user-imported 3D companions. It integrates on-device Apple Foundation Models for autonomous moods and a custom Swift/MLX port of a local decision model, showcasing advanced AI integration on Apple platforms.

Recent titles

1. Video deconstructing how Jev works
   2026-10-31 · youtube.com · HN discussion
2. What it costs to moderate 500 daily doodles with Jev
   2026-10-30 · johalloran.dev · HN discussion
3. Torturing Jev
   2026-10-30 · jonathanchiang.ca · HN discussion
4. Create your own Jev using Voyage models
   2026-10-28 · blog.voyageai.com · HN discussion
5. Jev and the Art of Truth Maintenance
   2026-10-26 · daniel.du-pless.is · HN discussion
6. mongo-predict: create your own Jev
   2026-10-26 · github.com · HN discussion
7. Everything on Wikipedia ranked pairwise by Jev
   2026-10-26 · jevipedia.com · HN discussion
8. The Rise and Fall of Jev
   2026-10-23 · blog.alexewerlof.com · HN discussion
9. Building a cheaper review loop for coding agents using Jev
   2026-10-23 · oxtan-maker.github.io · HN discussion
10. Jeffan-31B: A Gemma fine-tune ranked #1 on the Jev Decision Index
    2026-10-22 · huggingface.co · HN discussion
11. Smallchat (Jev as Judge) on the XERJ Corpus Hub
    2026-10-21 · hub.xerj.org · HN discussion
12. Show HN: Jev drives a browser – HN login and search in 8s for half a cent
    2026-10-21 · jev-browser.rovira.sh · HN discussion
13. Jeventus: A Jev powered Slack pet
    2026-10-20 · twitter.com · HN discussion
14. Testing Jev, a fast multiple-choice AI model, on forensic accounting tasks
    2026-10-20 · case-trail.com · HN discussion
15. Jev models make agents faster. Hooked control planes steal the difference
    2026-10-20 · pentad.ai · HN discussion
16. Show HN: Jev Cards, Can you beat a vision model at a picture quiz?
    2026-10-20 · jevcards.com · HN discussion
17. Jev is NOT a reranker (but wow it's close)
    2026-10-20 · kapa.ai · HN discussion
18. Jev wins against Clef and OpenAI Decisions
    2026-10-20 · latencyradar.com · HN discussion
19. Jev Helps Multiplayer AI
    2026-10-19 · runtype.com · HN discussion
20. Where Jev Fits in a Large-Scale Data Pipeline
    2026-10-19 · trychannel3.com · HN discussion
21. Ranking a Backlog with Jev
    2026-10-19 · feniix-hq.net · HN discussion
22. Show HN: Jev in action – task classifier and auto model router in Keen Code
    2026-10-19 · mochow13.github.io · HN discussion
23. Jev for Every Slack Message
    2026-10-19 · slacklookout.com · HN discussion
24. Eval of 19 ways to crop images using LLMs, Jev and image classification models
    2026-10-18 · workflows-of-ai.com · HN discussion
25. TypeSafe's Jev more susceptible to attack than LLMs, says security vendor
    2026-10-16 · infoworld.com · HN discussion
26. Testing Jev and other decision models on search-related tasks
    2026-10-15 · parallel.ai · HN discussion
27. How to Use Jev with Claude Code, Codex and Cursor
    2026-10-15 · flaviocopes.com · HN discussion
28. JEV: Type Safe AI
    2026-10-15 · typesafe.ai · HN discussion
29. Jev Meets the Trolley Problem
    2026-10-15 · zachlandes.com · HN discussion
30. Jev 1.14.0 is still bluffing on hard questions
    2026-10-15 · brussle.com · HN discussion

All trends