Entries 529 to 576 of 743, in the hub's default order. The introduction is on page 1.
Showing 529 to 576 of 743
A video demo classifies 100 test emails, 50 legitimate and 50 fraudulent, with Jev in 1.42 seconds, routing the 31 cases below 95% confidence to Kimi K3 for a second opinion. The combined pipeline reached 96% accuracy for about $0.07 total.
Hassan (@nutlope)OpenEmail fraud detection with Jev and Kimi (X) on x.com60k viewsJev pass 100 emails, 1.42sCombined accuracy 96/100
OpenRouter benchmarked Jev with its Ori Eval judging harness against popular LLMs and found Jev over 5 times faster than the next fastest model, with its slowest requests still beating every other model's median.
OpenRouter (@OpenRouter)OpenOpenRouter benchmarks Jev's speed against other models (X) on x.com60k viewsSpeed vs next model >5x fasterJev median latency 154 ms
A video demo of a prototype that feeds a live clinical consultation transcript to Jev, which traverses a medical ontology to classify symptoms, update the differential diagnosis, suggest treatments, and flag red flags in real time.
Chris McMaster (@rheum_ai)57k viewsOpenLive clinical-transcript classification prototype (X) on x.com
A live viral-post analyzer built with Jev: it scores viral potential and categorizes a tweet as the user stops typing, shown here as an early, in-progress demo.
A single prompt pulls about 100 reference images from Cosmos, NASA and The Met using Jev, shown as a video demo of the tool.
A video demo of a TSLA 50-day moving-average trading rule with Jev in the judgment layer: real API calls return typed answers, and the author's code decides the action from there.
A short video demo of an agent whose model router is powered by Jev; the post text gives no further detail on how the routing itself works.
A video demo of jev-review, an experimental local-first MCP plugin that scores a coding agent's work against quality metrics; the agent calls Jev, gets scored, makes improvements, and repeats the loop.
A video demo tests Jev as a real-time control policy for a robot arm in MuJoCo, splitting each update into two calls: one to decide the next action, another to decide how to move the arm and gripper. Jev receives simplified geometry and contact data as text rather than images.
Artificial Analysis scores 11 hosted decision models on knowledge, reasoning, tool-calling and safety tests, with cost and time per task, and runs language models on the same tests. Jev 1.13 leads the decision models at 58; Claude Haiku 5.5 (Low) scores 61 but takes 1.4 seconds per task against 0.13.
Artificial Analysis (@ArtificialAnlys)45k viewsOpenArtificial Analysis Decision Index: 11 hosted decision models against language models on artificialanalysis.ai
A leaderboard that runs Jev and 54 open System One models and clones through the same 120,000-question suite on one RTX PRO 6000, scoring accuracy and calibration. Jev leads the 0.2 edition at 51.67 on a chance-corrected scale, with AutoJev-27B close behind at 50.94.
apolinario (poli) (@multimodalart)43k viewsOpenJev Decision Index: Jev against 50+ open decision models on huggingface.co
A video demo shows background auto-fill buttons built with Jev integrated directly into an app's UI, contrasted with slower LLM-based alternatives.
A video demo of a logic interpreter built to compensate for the fact Jev cannot reason on its own: it runs a logic language whose facts and rules are written in plain English.
Open-weight decision model built on mmBERT-small that picks one of 2 to 20 supplied options on a CPU, with Apache 2.0 weights on Hugging Face. The launch page scores it against Jev reference values from a published protocol, beating them on three of four tasks and losing on the 72-label Banking77 test.
Supersonic Labs (@supersonicai)OpenJulia 1: a 144M-parameter decision model from Supersonic Labs on supersoniclabs.ia.br39k viewsParameters 144.3MGPU spend ~US$104
A video demo pits Jev against GLM 5.3 at chess: GLM won by checkmate in 29 moves, with Jev responding in about 0.3 seconds per move versus GLM's 5.8 seconds, for a total game cost of 24 cents.
Hassan (@nutlope)OpenJev vs. GLM at chess (X) on x.com39k viewsJev per move ~0.3sGame cost 24 cents
Jev picks the target for the MOSS litter-picking robot, for example switching from cans to bottles on a new instruction. In the newer v0.2 physics replays it also picks each arm action, and the page replays MOSS acting on Jev's real recorded decisions.
Runs Jev through support routing, refund detection, a prompt-injection attempt, exact-value selection, agent auditing and browser automation, with evaluation times as low as 92ms. Notes constrained outputs do not guarantee correct answers when the option list is incomplete.
Paste up to 10,000 words and see the generated-sounding sentences worth a second look. Free and no login.
Jozef (@jozef_gherman)OpenJev Detector: Free AI Slop Detector on jevdetector.com36k viewsScan speed ~2s/10k words
Argues that the purely engineering decisions inside security workflows belong to a typed decision model like Jev rather than a chat model.
Kostas Tsale (@Kostastsale)OpenRethinking security engineering with Jev (X) on x.com34k viewsSpeed and cost x200
A video demo combines OCR with Jev to build an image classifier that categorized about 900 images in 40 seconds.
Fayaz Ahmed (@fayazara)OpenOCR-powered image classification (X) on x.com34k viewsThroughput 900 in 40s
A demo of a general-purpose browser skill where an agent is given a site and a task, and Jev decides every click based on what is on screen.
AI Builder Club (@aibuilderclub_)33k viewsOpenJev-browser, a general browser skill for agents (X) on x.com
Hands-on test of Jev through the TypeSafe console with commentary on where it fits. 17 minutes.
A video demo of a per-page PDF router built with Jev that decides which pages need OCR and extracts the rest locally, aiming to cut both cost and processing time.
A video demo runs 6502 CPU emulation directly on Jev, executing short programs in real time with no speed-up, though the output occasionally diverges from a reference emulator.
A video demo rebuilds a job crawler with Jev: the model finds a company's careers page, chooses which links to follow, recognizes job postings, and scores each role against the user's profile, cutting the task from about 5 minutes with an LLM to about 20 seconds.
25k viewsBefore (LLM) ~5 minAfter (Jev) ~20s
30-second Snake benchmark: a local 421M Laya checkpoint scores 46 at 86.5 decisions per second, while Jev 1.13.0 over the cloud API scores 1 at 3.2 decisions per second.
Simplifying AI (@simplifyinAI)OpenLocal Laya vs Cloud Jev on the Same Snake Game on x.com24k viewsLaya speed 86.5/sJev speed 3.2/s
An experimental linter for Postgres Row Level Security policies, built with Jev. It sends nine judgments about a submitted policy and table schema in a single request to check whether access controls actually restrict users to their own data, not just that RLS is turned on.
A Clippy-style assistant that watches product usage and only appears when Jev judges the user is hesitating, confused or stuck; Jev also drives its reactions.
Breaks down the side-by-side against GPT-5.6 Terra, the 193.6x faster and 444.6x cheaper workflow evals, and the 0% type-error chart. Covers the Doom and Wikiracing demos.
Given a dataset of outreach messages, Jev found which intent signals correlated with the most booked demos in 40 seconds for under $0.20; the poster says it can also rank leads and measure prospect-message fit.
Pierre-Eliott Lallemant (@pierreeliottlal)OpenFinding the intent signals that book demos with Jev (X) on x.com22k viewsTurnaround 40sCost <$0.20
A video demo of Jev tagging X posts in real time as breaking, a golden nugget, or slop.
Save notes, links, images, videos, PDFs, audio, and other useful material as clean Markdown your AI agents can search, read, and update.
Ian Nuttall (@iannuttall)OpenKeep anything as Markdown for your AI agents on keep.md20k viewsSearch rerank 7x fasterTagging speed 50x faster
A video demo uses Jev to filter out low-quality reply comments on X that the platform's own filters miss, built in about 5 minutes with Astra and an API key.
Ian Nuttall (@iannuttall)OpenFiltering unwanted X replies (X) on x.com18k viewsBuild time ~5 min
Examines the speed and cost claims and what the published benchmarks show, noting Jev competes on some tasks and falls well behind on others. Clarifies the zero-hallucination claim covers output structure, not decision correctness.
A video demo shows Jev-thoven, a system built on Jev, composing an entire two-minute song from first principles.
Explains the autoregressive latency floor set by memory bandwidth per token and how a single-pass architecture bypasses it. Covers the Kahneman framing, calibration and the Jevons Paradox naming.
A video demo of controlling Figma with voice commands routed through Jev, with a total cost of $0.01 from first line of code to the recorded demo.
A video demo of Bannerbear using Jev to map template fields such as photo, name, and company name to differently named data source fields such as avatar, full name, and business, in one click.
A video demo of an app that monitors the macOS clipboard with Jev, detecting when the copied content is a terminal command and offering a quick action to run it.
A live trading site where Jev re-reads a Solana token every 5 seconds for its call, dump risk and buyer control. Users write their own yes-or-no, choice or score classifiers, combine up to six into a strategy, and backtest or run it live against every new launch.
Short introduction to Jev focused on RLCD, Reinforcement Learning for Calibrated Decisions, as the training method behind the model.
A demo of a tool that uses Jev to automatically find a website's billing page, list and download its invoices in one click, and remember where invoices live for next time.
On OpenMausBot, Jev reads an incoming task and decides which bot to wake and which underlying LLM to assign it, working across the user's existing LLM subscriptions.
A video demo of Jev, now available through the Venice API, classifying 24,000 Hacker News posts into 12 categories in about 2 minutes.
Sabrina (@sabrinaesaquino)14k viewsOpenClassifying 24,000 Hacker News posts on the Venice API (X) on x.com
A maze demo comparing decision latency: local Laya on a base M5 responds at a 19ms median, versus roughly 300ms from Jev's cloud API for the same task.
A demo of the Clonk agent orchestrator launching agents on a canvas, using Jev instead of waiting on a slower general-purpose model loop for each step.
Livestream coding session against the Jev API, plus discussion of network architectures that could produce this kind of JSON predictor model.
Podcast segment covering the launch, the $40M raise, Almeida's path from RLHF to RLCD, and the 70-500ms response times. Explains the name as a nod to Jevons Paradox.
Nothing matches those filters
Clear a filter to widen the search, or send us the thing you were looking for.
