Updated
Desk
Timeline
71 stories grouped by the date in the article, newest first. That date is the primary post or write-up, not the day this desk rewrote it. For how a story gets on the list, see method.
15
AgentOS posts a 121-turn routing table for an opt-in Jev strategy
AgentOS added Jev as a third Pilot Router strategy, off by default. On the project's 121-turn labeled set the chart shows Jev at accuracy 0.901 against the local MiniLM classifier at 0.355. The same chart prices 121 cases at about $0.005 and a call at about 0.9 seconds. A follow-up pull request the same day says it sharpens the R2 and R3 criteria.
Exa posts a demo that puts web results into Jev's state
Ishan Goswami posted a demo at demos.exa.ai/jev-web-context. The post says Jev without web search is confidently wrong, and Jev with Exa results is more accurate. The page we loaded lists seven yes/no prompts and does not print a before-and-after. Diogo Almeida quoted the post and asked if it was real.
jev-gateway asks Jev which tool a coding agent should call
Vini Lana posted jev-gateway, an MIT local proxy for Codex, Claude Code, OpenCode, and Gemini CLI. Jev picks the tool. Closed arguments can skip the language model. On the project's chess-engine bench, GPT-6 Astra and GPT-5.6 Sol cut bug-fix output tokens by 57%. The feature task made Opus 5 and Sonnet 5 use more.
jev-rubiks asks Jev when to speak, after a test where picking moves did nothing
maxlibin/jev-rubiks is an MIT browser cube. Kociemba solves it in code, in about 20 ms. Jev only decides whether the coach should talk. The README says a removed experiment let Jev pick among 18 turns, forty times, and the scramble distance stayed at 22, the same as a random walk that went from 22 to 21.
jym asks Jev which real subcommand a mistyped one was aiming at
syumai posted jevyoumean, an MIT Go wrapper. When a subcommand is not in the CLI's help, it sends one Jev Choice over the documented names and descriptions, then offers the top matches. Valid commands never call the API. The README ships an eval harness and does not print a scored corpus.
jsort ranks a file by asking Jev which of two texts is more of something
Khaled Eltokhy published jsort, an MIT command that fits a Bradley-Terry scale from pairwise Jev judgments. On 300 CommonLit excerpts it tracked teacher easiness at Pearson r 0.824 for about five cents. Joe Weisenthal then said he had rescored 4,005 FOMC speeches and statements with it. The repo's own Fed check is 95 opening statements.
laya-mlx runs the Laya checkpoints on Apple Silicon
mizorewww/laya-mlx is an Apache 2.0 MLX port of Convai's Laya weights. On an M3 Max the README's one-question median is 13.42 ms for the 421M checkpoint and 7.39 ms for the 322M multilingual checkpoint. Selected answers matched upstream on 63 of 63 validation questions. The port does not publish a new accuracy table against Jev.
llamacpp-jev answers /v1/systemone from an unmodified llama-server, including images
Chirag posted llamacpp-jev, an MIT wrapper that puts TypeSafe's decision API in front of stock llama-server. On an M4 Pro, four questions about a fresh 448 by 448 image took 526 ms median, 32 of 32 on synthetic shapes. The probabilities are a raw label softmax. Qwen3.5-2B picks option 16 on every 64-way question the notes tried.
TypeSafe pauses new Jev signups
TypeSafe posted at 06:19 UTC on September 22 that new signups are paused while existing accounts keep working. Diogo Almeida said the service is overflowing and that the pause lets the team sleep. The post does not say when signups reopen.
A 612-row banking-feed test puts Jev under two Gemini rows on accuracy
Peter posted a categorization run on 612 sanitized rows from a private banking-feed set of about 33,000. Jev 1.13 scored 36.1% in 6 seconds at $0.023. Gemini 2.5 Flash scored 37.3% in 95 seconds. Gemini 3.7 Flash scored 42.5% in 78 seconds. The post names no repository and no category list.
meraGPT posts Decider 1 ahead of Jev on a teacher-labelled set
Latent Node posted Decider 1 (sd-1), a hosted model that answers noul, choice, and score questions in one call. The blog's table is accuracy 0.768 against Jev 1.13.0 at 0.727. The labels are an average of teacher samples. The Hugging Face viewer lists 1,600 rows. The blog says 400 cases and 2,000 decisions.
dejevu books the same Flights search faster than the published Jev demo, on another machine
Idov Mamane posted dejevu, an MIT browser loop that sends one JSON action to an ordinary chat model. On the README's table, llama-3.3-70b on Groq finishes the Flights check in 5.63 seconds against jev-ultrafast's published 7.09 seconds. The Jev column was not rerun here. That Groq route costs more per run.
jevelry turns a Jev call into a folder, a verdict, and a log
Pavel Hegler posted backant-io/jevelry, an MIT npm runtime. A jevel is a folder with one JEVEL.md. The CLI asks Jev once, maps each answer to act, mark, or fall_back, and appends a log. The live sample in the README is one state on jev-1.13.0. Three on-demand API calls took 2.49 seconds.
pg-jev puts prompt_jev() in PostgreSQL and returns jsonb
Jose Mejias posted mejiasd3v/pg-jev, an MIT extension. prompt_jev sends text to Jev from SQL and returns a probability, a choice, or a score as jsonb. The README we read has no accuracy or latency table. Calls are not batched across rows.
A Hold'em table asks Jev about the opponent and lets code pick the bet
A September 22 post links Jinzhengxu/poker-table and says each action averaged 545 ms and a hand cost $0.002. The README's Jev path is one round trip about the opponent, after a local equity table. A 46-spot bank, three repeats, is described at about one second median. Those two clocks are different measurements.
11
djev reads DiffusionGemma as a decision API, and sits third on JevBench
Florian S's JevBench table puts djev at 74.3, 1.1 behind Jev. Maisa's djev.dev hosts the API at an announced $0.035 per million input, with native image questions. Matt Mastracci posted phone-camera clips of DiffusionGemma-as-Jev. Davipar/djev-dev is the Apache 2.0 local stack.
A native DuckDB extension batches Jev over SQL rows at about 2,300 rows per second on a live run
Prasanth J posted duckdb-jev, an Apache 2.0 C++ extension that asks TypeSafe's Jev from SQL. A live Choice run classified 2,049 synthetic rows in 0.887 seconds. A cached replay of the same query finished in 28.6 ms with zero API calls. Row text leaves DuckDB.
jevsearch re-ranks keyword hits with Jev and scored 83% Hit@1 on the author's TypeSafe docs set
Kyle McLaren posted jevsearch, an MIT shadcn registry block that streams keyword hits and then re-ranks them with Jev. On 41 labelled queries over TypeSafe's documentation, Hit@1 rose from 41% on the keyword pass to 83% after Jev. Median uncached ranking was 278 ms. About $0.26 per thousand searches on that corpus.
MotherDuck's prompt_jev() labels 100,000 AG News rows in 40 seconds for $0.50
MotherDuck shipped prompt_jev(), a SQL function that calls TypeSafe's Jev. On 100,000 articles drawn from the AG News training split, the published table is 2,484 rows per second, 89% accuracy, $0.50, and 40 seconds. GPT-5.6 Terra on the same sample is 52 rows per second, 88%, $37.58, and 31 minutes 59 seconds.
An n8n community node routes Jev answers onto category branches, including Needs Review
khmuhtadin posted n8n-nodes-jev-classification, an MIT community node that wraps TypeSafe's Jev as classify, score, check, and ask. One output per category, plus a Needs Review branch under a confidence threshold. n8n declined to verify it because it is building a built-in classifier. Self-hosted only.
Open-Jev's released 9B checkpoint answers 179 of 231 public JevBench tasks
Zefan Cai published an audited run of the already released Open-Jev 2B and 9B checkpoints on the 231 public JevBench tasks. The 9B model is correct on 179. Hosted Jev 1.13.0 is correct on 200, GPT-5.6 Luna on 206, and GPT-6 Astra on all 231. The other 303 tasks stay private, so this is not a full JevBench score.
A 100-case security triage demo put Jev at 65.3% balanced accuracy against Terra 75.1% and Opus 91.2%
RZ posted a recorded comparison of Jev, OpenAI Terra in Codex, and Claude Opus in Claude Code on 100 synthetic TypeScript security cases, five passes each. Jev's balanced accuracy was 65.3% at 0.57 s median. Terra 75.1% at 10.23 s. Opus 91.2% at 13.83 s. Jev repeated the same answer on 95 of 100 cases. Cost about $0.015 versus $15.02 and $13.94 for 500 decisions.
SimpleJev turns an open model into a classifier by reading the next-token logits
Featherless posted SimpleJev, an Apache 2.0 server that scores Choice, Score, and Noul from a Hugging Face model's next-token logits. A public demo needs no key. JevBench listed Qwen3.8-27B at 67.3, with hard-tier accuracy 75.0%, matching Jev on that slice and trailing on the composite.
Spring publishes a Java client that asks Jev before the chat model runs
Christian Tzolov wrote up spring-ai-community/spring-ai-typesafe, an Apache 2.0 Java SDK and a set of Spring AI advisors. On his laptop a three-question ticket call took 310 ms. The blog says 0.1.0 is on Maven Central. The README quick start still shows 0.1.0-SNAPSHOT, and a search of search.maven.org returned no documents for those artifact ids.
TypeSafe posted an API and console outage, then said it was back
At 08:30 UTC on September 21, TypeSafe posted that the API and console had issues and pointed at status.typesafe.ai. At 09:53 UTC it posted "We are back!" The status incident is titled "Console is unavailable" and marked resolved at 08:16 UTC, before those tweets.
Vercel's form router keeps Jev's destination when confidence is at least 0.95
Vercel published a Next.js template that asks Jev to pick a team for a form submission and sends the same submission to GPT-5.6 Luna Fast when confidence is below 0.95, missing, or the call fails. The fallback does not see Jev's answer. Tests mock both models.
8
AstroHan ran 30 coding tasks with a Jev context filter: 25 passed, 22 without it
AstroHan scored every tool result with Jev before it reached deepseek-flash, on the same 30 FrontierHarness tasks. Filtered 25/30 pass@1 versus 22/30 unfiltered. On meriyah the unfiltered run timed out at 0/49 tests; the filtered run passed 49/49. One run per arm. Token spend did not fall.
DocJev classifies and splits real PDFs with Jev, 5.7x to 6.5x faster than Luna on the author's set
Jerry Liu posted DocJev, an Apache 2.0 library that classifies and splits PDFs with local LiteParse and hosted Jev. On 40 public-sector originals, both engines scored 40/40. Jev split 7 of 8 packets exactly against Luna 8/8, at 5.73x and 6.45x the median decision speed.
TypeSafe drops the Jev waitlist; new accounts start with $5
Five days after launch, TypeSafe posted that Jev is available to everyone, with no waitlist. Every account starts with $5 of credit, about 120 million input tokens at the published $0.042 per million.
jevctl puts Jev on the command line, with a compaction hook that keeps lines verbatim
jaidev posted jevctl, a friend's npm CLI that wraps TypeSafe's Jev as shell commands. verify, screen, classify, extract, find, rerank, match, route, ask, compact, and batch. Keys can be TypeSafe, OpenRouter, or Cloudflare. compact drops stale tool results the way Tran's sieve does, at a 0.5 keep threshold.
Jevflake puts Jev in Snowflake: noul, choice, and score as SQL functions
Luke Kranz published Jevflake, a dbt package (and a Terraform module) that lets Snowflake call TypeSafe's Jev API from SQL. Answers are stored once, reused, and tested like any other model. Version 0.1, run against one live account. Row content leaves Snowflake.
Jared Palmer's Kev family copies Jev's API on Qwen, and trails it on new sources
Cognition's VP of engineering published Kev, an Apache 2.0 family of Jev-like models you train and serve yourself. On data the models never trained on, Kev-8B scored 79.6% against Jev 85.7%. JevBench listed kev 0.6B at 66.7. A Qwen3.5 generation followed the same day.
Laya, an Apache 2.0 System One stack, publishes where it beats Jev's numbers and where it does not
Louis Grenard pointed at Laya, Convai Innovations' open typed-decision engine. Three checkpoints, a Router that picks among them, Apache 2.0 weights on Hugging Face. The author's table puts routed Laya above published Jev on some English sets and far below on Banking77. Jev was not called. JevBench's independent row is 70.1 against Jev 75.3.
semgate asks Jev about each HTTP request, then a live range to attack it
mizutani published semgate, a Go net/http middleware that sends each request to Jev as typed questions and lets your function block, route, or pass. A companion range, Injection Range, is up on Cloud Run. Default configuration sends every header, including Authorization and Cookie, unless you denylist them.
10
GitHub Next ships LocalJev, a prompted stand-in that speaks /v1/systemone
GitHub Next published LocalJev, a Bun server that turns Jev questions into JSON-probability prompts for models running on oMLX. A 1,200-request bake-off on an M5 Max compared five 4-bit checkpoints. The authors say the probabilities are self-reported, not read from logits.
LlamaIndex's Logan Markewich ships jeff, a local System One API on GLiFormer
Logan Markewich released jeff, a self-hosted drop-in for TypeSafe's /v1/systemone endpoint backed by Knowledgator's GLiFormer. On 1,600 public items it trails Jev, especially on irony and reading comprehension, and comes in cheaper to serve on a GPU.
Jev macOS Loop clicks native Mac windows; Finder sorted nine files in 7.39 seconds
Josh C. Simmons open-sourced a computer-use loop for Apple silicon that keeps pixels on the Mac and asks Jev only over observed controls. A Finder demo moved nine dummy files in three group drags in 7.39 seconds, with a separate verifier checking locations and hashes.
LiteLLM proxies Jev's evaluate endpoint and uses it to drop stale tool results
LiteLLM added a TypeSafe pass-through in v1.103.0-rc. Clients hit POST /typesafe/v1/systemone; spend is logged as typesafe/jev-1.13.0. The same proxy can ask Jev whether older tool results still matter, and replace those below 0.2 with a short notice before the generator runs.
JevBench scores typed decisions on four axes, and Jev's lead is a pricing story
Florian S published JevBench, a 534-decision suite for models that return a label and a probability. The composite is a geometric mean of intelligence, calibration, speed, and cost. Jev 1.13.0 sits at 75.3. A 4B open rebuild is 0.7 behind. GPT-5.6 Luna wins the hard items and loses the score on price.
Jevis is a Flutter integration-test agent that only taps what you registered
Jake Gwon released jevis, a Dart package that drives Flutter's integration_test loop with Jev. You register legal actions, write a goal, and set an action budget. Jev picks the next widget action and a Noul checks whether the goal is already on screen.
LangChain ran Jev as a judge: 500 matching pass/fail labels, and a $0.34 bill against Claude's $28
Daniel Shea and Seán Roche compared Jev with GPT-5.6 Luna, GPT-5.6 Terra, and Claude Sonnet 4.6 on five frozen weather-agent traces. Jev matched a human pass/fail label on all 500 repeated decisions. Continuous quality-score variance was 92 to 913 times lower than the LLM judges. Five cases, one reviewer.
Jason Lemkin tried Jev on SaaStr Connect and stopped on a 32% false-admission rate
Jason Lemkin ran 600 judgments against Claude Sonnet. Jev cost $0.064 against about $5, roughly 90 times cheaper. A blind third-model referee scored Jev 70.5% and Sonnet 77.5% on 200 random pairs. Jev let through 47 of 148 negatives. He says that is too high for Connect.
SemIf, a 4B open rebuild that reads option logits, sits 0.7 behind Jev on JevBench
Theodore Lee's SemIf (formerly OpenJev) scores typed options from a frozen Qwen3.5-4B by reading logits, not by sampling JSON. On JevBench v1.2.2 it is 74.6 against Jev 1.13.0 at 75.3. The author's own 3090 timings put direct readout at 1.023 s for 21 questions against 5.332 s for a compact JSON array.
Sam Saffron tries Jev as a hidden control plane for GPT Live in term-llm
Discourse cofounder Sam Saffron opened a term-llm pull request that classifies each GPT Live voice turn with TypeSafe's System One API in about 650 milliseconds, instead of a tool-calling fast-model turn. Shadow mode is on by default. Failures pass the utterance through unchanged.
13
Aman Kumar ran about 16,000 Jev calls and called it a filter, not a classifier
Aman Kumar compared Jev with gpt-5.4-mini and gpt-5.6-luna on four public sets, then on production page gates and an inbox. On short text it tied or led. On a whole-page read it fell apart. He published the public cases and per-item answers.
Bespoke Nimble is an open 9B Jev-style model: 90% on its own 324-example holdout
Bespoke Labs released Nimble, a Qwen3.5-9B LoRA trained on contrastive synthetic pairs, plus the data and recipe. On 324 held-out examples it matched 90.12% of the reference labels. Jev 1.13.0 matched 93.21%. The base 9B model matched 66.36%.
Charly Poly put Jev next to open encoders, not GPT-5.6, and it lost on calibration
Browserbase's Charly Poly ran Banking77 against ModernBERT, BART-MNLI, and GLiClass. Jev won macro-F1 and automated about twice as much traffic at a 90% precision bar. Expected Calibration Error favored the 396M open encoder, and Jev assigned probability 0 to the right label 6.6% of the time.
Hono's author routes HTTP by meaning, then warns you not to
Yusuke Wada published hono-jev-router, an experimental Hono router that asks Jev which plain-language description matches an incoming request. The README treats it as a demo with sharp limits.
TypeSafe lists nine ways Jev 1.13 is still jagged, and asks for more
Cofounder Sasha Sheng posted that Jev is still too slow, too expensive, and too dumb, and pointed at TypeSafe's jaggedness page for jev-1.13. The docs, last reviewed September 17, name nine failure modes and ask people to file more in Discord.
Jev scored 384 morning headlines in 25 seconds; Opus 5 finished four
Elvis ran a morning Google News wire through Jev and Claude Opus 5 at the same time. Jev tagged 384 headlines for 15 brand desks in 24.9 seconds at $0.19. Opus got through four headlines and cost $0.77 before the run was stopped.
OpenRouter, Cloudflare, and Netlify start serving Jev
Three days after launch, Jev is callable through OpenRouter's beta Decisions API, Cloudflare Workers AI, and Netlify Functions. The TypeSafe waitlist is no longer the only door.
Venice lists Jev on its API: same key, no copy of the state
Venice put Jev on a beta Decisions API the night of September 18. Calls go to POST /api/v1/decisions with model jev-latest, billed at TypeSafe's $0.042 per million input tokens with free output. Venice says it keeps no copy of the state or the answers, and TypeSafe quoted the launch.
A hybrid Jev overlay sped up Postgres joins 12% on JOB, and made the planner slower
Michael Malis asked Jev to help plan Postgres joins on the join order benchmark. Letting it pick the join order made queries twice as slow. Estimating filters, then overriding Postgres only when Jev was confident, improved geometric-mean time by 12%.
SREGym put Jev on a Codex agent: 24 passes out of 50, up from 20
Jackson Clark's team wired Jev into SREGym-Lite as a test ranker and evidence reviewer on a Codex agent that still diagnosed and repaired. With gpt-5.6-luna, the Jev-assisted agent passed 24 of 50 attempts against 20 without it. Two problems got worse.
Almeida's bet on Jevons paradox over chatbot price cuts
In replies three days after launch, TypeSafe's founder argued that the business depends on enabling new software loops rather than capturing existing language model budgets.
TechCrunch covers Jev, and quotes a Vercel engineer who swapped Luna for it
Tim Fernholz's TechCrunch story is the first large press write-up of Jev. It reports that TypeSafe briefly could not serve the API under demand, quotes Pranit Sharma at Vercel on a 5 to 18 times speedup versus gpt-5.6-luna, and points at Nikhil Mudholkar's email test against Gemini.
Vercel says Jev is the fastest-adopted model in AI Gateway history, and makes it free until Sept 25
A Vercel blog post dated September 18 reports that Jev reached nearly 13% of paid AI Gateway teams in its first 24 hours, twice the GPT-5.6 family and more than six times Fable 5.1. The model card and a September 19 Developers post say promotional pricing runs through September 25, 2026.
8
A week of Jev repos: routers, MCP, graphs, games
A widely shared Awesome Jev thread listed fourteen working patterns; community directories passed a hundred projects within days. The repeating pattern is choose, score, and stop.
Bryo ran 1,565 business emails: Jev at 96.4%, Gemini a point or two higher
Nikhil Mudholkar, CTO of Bryo AI, routed 1,565 German and English supplier emails into 10 categories. Jev scored 96.4% against Gemini 3.5 Flash-Lite at 97.5% and Gemini 3.8 Flash at 98.5%. Jev was about 10 and 22 times cheaper per thousand emails, and its high-confidence band had no errors on this set.
Browser Use wires Jev into the click loop: 7 seconds, $0.0039
Gregor Zunic showed a 1x-speed Google Flights run where Jev picks the action and DOM target each step, and a small LLM types only when it has to.
Compaction without summarization: score the tool calls and drop them
Tamara Tran argued that agent context compaction is still a summarization prompt in 2026, and that Jev can instead keep or drop tool output verbatim.
LangChain ships a Jev classifier, then parks it on the agent loop
Official LangChain docs add langchain-typesafe, introducing a Runnable for classification alongside experimental middleware hooks for agent loops.
Stagehand, Chrome, and a11y trees: computer use as a Jev question
Kyle Jeong of Browserbase and other builders structured computer-use around structured observation, typed queries, and automated execution, reporting sub-cent task costs.
Tetris as a legal-move test: Jev picks from the landings you give it
Two independent demonstrations evaluated Jev playing Tetris from enumerated legal placements, comparing latency and costs against Claude Haiku 4.5 and Gemini 3.5 Flash-Lite.
TypeSafe starts pulling people off the Jev waitlist
TypeSafe began admitting waitlist users in batches, reporting 20,000 approvals in an hour while addressing browser compatibility issues.
3
One builder ran Jev on 1,500 of his own emails
Ryan Vogel posted a long clip of Jev classifying his mail. The sample size is real; the accuracy number is not in the post. Parallel calls were the part viewers argued about.
A buy-or-sell every Monad block, with a mock default model
Jarrod Watts demonstrated a trading bot querying Jev for buy or sell decisions on each Monad block. The open-source repository runs in dry-run mode by default.
JevPilot is a playable driving sim that asks Jev which path to take
Justin Schroeder wired Jev into a Three.js driving simulator in under an hour, then open-sourced JevPilot the next day. The model sees compact tables of candidate paths, not pixels. Requests run up to four times a second in traffic.
3
Every ran 777 judgments in under a second and still wanted a better accuracy check
Mike Taylor at Every put 37 documents through 21 AI-writing questions at once. Jev returned 777 answers in under 0.7 seconds for about a quarter of a cent. A later defect test with Dan Shipper missed one of seven planted problems that Claude Fable 5.1 caught.
TypeSafe leaves stealth and ships Jev, a model that refuses to chat
Former OpenAI researcher Diogo Almeida announced Jev, a System One model built for typed software decisions rather than text generation, following two years in stealth.
What Jev returns: System One in plain language
TypeSafe defines a System One model as parallel, typed, and calibrated decisions without text generation, distinguishing it from language models operating in JSON mode.