Updated

Glossary

Terms

171 definitions from the Terms block on each article. A word can appear in more than one story. The wording here follows the explainer when we have one, otherwise the longer note.

__none__
A required option in jym's Choice, alongside the real subcommands. If Jev picks it, the wrapper does not substitute a documented command.

jym asks Jev which real subcommand a mistyped one was aiming at

2026-09-19-public-v5
The frozen recorded run on RZ's dashboard. One hundred synthetic cases, five passes, 1,500 decisions. Earlier v2 through v4 directories are rejected by corpus hash.

A 100-case security triage demo put Jev at 65.3% balanced accuracy against Terra 75.1% and Opus 91.2%

Abort-on-finish
The Opus 5 column is stopped when Jev completes the wire, so the slower model is not billed for the remaining headlines.

Jev scored 384 morning headlines in 25 seconds; Opus 5 finished four

Accessibility tree
The structural snapshot of a webpage exposing element roles, labels, and states. These workflows supply this tree to Jev as state rather than transmitting screenshot images.

Stagehand, Chrome, and a11y trees: computer use as a Jev question

Accurate when confident
In this write-up, accuracy where a pick-one confidence is 0.9 or higher, or a yes/no probability is at or below 0.1 or at or above 0.9, plus the share of items in that band.

Aman Kumar ran about 16,000 Jev calls and called it a filter, not a classifier

Action budget
The attempts argument on test(). It is a cap on UI actions executed, not on scenario restarts. Goal evaluation also runs before the first action and after the last allowed action.

Jevis is a Flutter integration-test agent that only taps what you registered

Action space
The set of valid moves available for the active piece, recomputed on each turn similarly to how browser agents rebuild interactive DOM elements.

Browser Use wires Jev into the click loop: 7 seconds, $0.0039 Tetris as a legal-move test: Jev picks from the landings you give it

Action table
The compact state Jev sees: eligible paths, road boundaries, nearby traffic, signals, stop memory, and destination guidance. Shared values are sent once.

JevPilot is a playable driving sim that asks Jev which path to take

AG News training sample
MotherDuck's published prompt_jev table uses 100,000 rows drawn with a reservoir sample, seed 43, from the AG News training parquet. It is not the dataset's held-out test split.

MotherDuck's prompt_jev() labels 100,000 AG News rows in 40 seconds for $0.50

AI Gateway
A cloud provider's managed proxy for AI APIs handling credentials, billing, and logs, such as those provided by Vercel, Cloudflare, and Netlify.

OpenRouter, Cloudflare, and Netlify start serving Jev Stagehand, Chrome, and a11y trees: computer use as a Jev question

anchored scale
A jsort JSON file of anchor texts and their scores. Later texts are compared only with those anchors, so a new line can be placed without sorting the original file again.

jsort ranks a file by asking Jev which of two texts is more of something

AutoModeMiddleware
An experimental wrap_tool_call hook where Jev evaluates tool call safety, returning an error ToolMessage when an action exceeds risk thresholds.

LangChain ships a Jev classifier, then parks it on the agent loop

Awesome list
A curated catalog of projects. Two independent lists appeared during launch week; project counts on each reflect point-in-time snapshots.

A week of Jev repos: routers, MCP, graphs, games

Balanced accuracy
Unweighted mean of per-class recall over vulnerable, safe, and insufficient context. The published v5 run puts Jev at 65.3%, Terra at 75.1%, and Opus at 91.2%.

A 100-case security triage demo put Jev at 65.3% balanced accuracy against Terra 75.1% and Opus 91.2%

banking feed
In this post, rows from a product that categorizes bank transactions. The 612 rows are described as a sanitized, harder subset of an evaluation set of about 33,000. The categories are not listed.

A 612-row banking-feed test puts Jev under two Gemini rows on accuracy

Borderline-weighted pairs
Lemkin's 200-pair slice that over-represents close cases. He reports 59.5% agreement with Sonnet there, against 67% on 200 random pairs.

Jason Lemkin tried Jev on SaaStr Connect and stopped on a 32% false-admission rate

Calibrated decision
TypeSafe's phrase, used in the TechCrunch piece, for a typed answer with a probability rather than generated text.

TechCrunch covers Jev, and quotes a Vercel engineer who swapped Luna for it

Calibration
The degree to which a model's reported probability corresponds to empirical label accuracy, which TypeSafe targets via RLCD.

What Jev returns: System One in plain language

Choice
Selecting one option from a user-defined set (cardinality up to 255), returning a probability distribution and confidence score.

One builder ran Jev on 1,500 of his own emails What Jev returns: System One in plain language

classifier.dev fast tier
A hosted classification API whose own page, and JevBench's note, say the fast tier is Jev. It matches Jev on Intelligence and leads the ranking on a cheaper published plan and a faster measured p50 from the bench's server.

JevBench scores typed decisions on four axes, and Jev's lead is a pricing story

coach facts
In jev-rubiks, the plain-English description code sends after a hand move: beginner stage, pieces placed, whether the move undid progress, and how long the coach has been quiet. Jev answers whether to speak. It does not return the next algorithm.

jev-rubiks asks Jev when to speak, after a test where picking moves did nothing

Compaction
Reducing an agent's context history so subsequent prompts fit model limits. Traditional approaches rely on summarization prompts; this method scores entries and removes irrelevant ones.

Compaction without summarization: score the tool calls and drop them

Confidence band
A slice of answers above a probability cutoff. On this set he reports zero errors in Jev's most confident 85.5%, and that nearly half of answers below 70% were wrong.

Bryo ran 1,565 business emails: Jev at 96.4%, Gemini a point or two higher

confidence statistic
TypeSafe metadata on a Jev Choice, keyed by question id. The form router reads it from providerMetadata.typesafe.confidence. It runs from 0 when probability is spread across destinations to 1 when it sits on one destination, and it is not the probability of the selected option.

Vercel's form router keeps Jev's destination when confidence is at least 0.95

Context filter
AstroHan's loop: Jev answers one boolean per ~2k-character tool chunk (state is task, intent, action, output) and the harness keeps the chunk when p > 0.5.

AstroHan ran 30 coding tasks with a Jev context filter: 25 passed, 22 without it

Contrastive data curation
Write two examples that differ in one fact so the correct answer flips. The rest of the context and the question stay the same.

Bespoke Nimble is an open 9B Jev-style model: 90% on its own 324-example holdout

Control plane
In this PR, a classifier lane that labels each GPT Live utterance before the main model runs. Config values are off, agent, or classify.

Sam Saffron tries Jev as a hidden control plane for GPT Live in term-llm

Decider 1
meraGPT's hosted model sd-1. The September 22 blog says it answers noul, choice, and score questions in one call, and that a typesafe-sdk client can point TYPESAFE_BASE_URL at https://meragpt.com.

meraGPT posts Decider 1 ahead of Jev on a teacher-labelled set

Decision rate
How often the sim asks Jev for a new path. The README caps this at 4 Hz in traffic and about 1.5 Hz on an empty road. Single-answer questions are resolved locally.

JevPilot is a playable driving sim that asks Jev which path to take

Decisions API
Venice's endpoint POST /api/v1/decisions, which accepts state and typed questions rather than a chat transcript. OpenRouter uses the same name for POST /api/alpha/decisions.

OpenRouter, Cloudflare, and Netlify start serving Jev Venice lists Jev on its API: same key, no copy of the state

Direct logit readout
One forward pass over declared option tokens, with no sampled answer. SemIf's 3090 note: 1.023 s for 21 questions versus 5.332 s for a compact JSON array from the same weights.

SemIf, a 4B open rebuild that reads option logits, sits 0.7 behind Jev on JevBench

direct mode
A jev-gateway outcome. Jev is confident about the tool, and every argument is an enum, a boolean, or a constant, so the gateway writes the tool call and does not call the language model.

jev-gateway asks Jev which tool a coding agent should call

djev
Maisa's hosted decision API on DiffusionGemma, and the name Matt Mastracci used for DiffusionGemma-as-Jev. JevBench v1.2.8 lists the hosted row at 74.3.

djev reads DiffusionGemma as a decision API, and sits third on JevBench

DocJev
Jerry Liu's Apache 2.0 library for classifying and splitting PDF, DOCX, and PPTX files. LiteParse reads pages locally; hosted Jev returns the category or the page-range boundaries.

DocJev classifies and splits real PDFs with Jev, 5.7x to 6.5x faster than Luna on the author's set

does_pass
The binary evaluator in jev-as-a-judge. Jev matched the human oracle on all 500 repeated decisions in this run.

LangChain ran Jev as a judge: 500 matching pass/fail labels, and a $0.34 bill against Claude's $28

Dry run
Operating against live market data without private transaction keys, simulating order fills when trade prints cross the resting quote.

A buy-or-sell every Monad block, with a mock default model

duckdb-jev
Prasanth J's Apache 2.0 native C++ DuckDB extension. SQL functions call TypeSafe's Jev over HTTP, with batching, streaming, and an optional connection cache.

A native DuckDB extension batches Jev over SQL rows at about 2,300 rows per second on a live run

Early warning
Taylor's description of using Jev as a first pass that is cheap enough to run even when a later human or larger model still reviews the result.

Every ran 777 judgments in under a second and still wanted a better accuracy check

evaluate
Vercel AI SDK's experimental method for Jev, passing state and questions to receive typed answers rather than text completions.

TypeSafe starts pulling people off the Jev waitlist

Expected Calibration Error
A score for how far stated probabilities sit from observed accuracy. Lower is better. Poly reports 0.081 for ModernBERT and 0.105 for Jev on this Banking77 run.

Charly Poly put Jev next to open encoders, not GPT-5.6, and it lost on calibration

Fail open
Every error path returns the original input as steer, so a classifier outage does not drop the utterance.

Sam Saffron tries Jev as a hidden control plane for GPT Live in term-llm

fail_open
LiteLLM's default when Jev is unavailable: send the original request to the generator without compaction.

LiteLLM proxies Jev's evaluate endpoint and uses it to drop stale tool results

fallbackReason
The form router's label for why Luna decided. null when Jev cleared 0.95. low-confidence, missing-confidence, or jev-error otherwise. A failed Luna call does not return a reason or a destination.

Vercel's form router keeps Jev's destination when confidence is at least 0.95

False admission
A negative case the model treats as acceptable. Lemkin reports 47 of 148 for Jev (32%) and 20 for Sonnet (13.5%) on this SaaStr Connect sample.

Jason Lemkin tried Jev on SaaStr Connect and stopped on a 32% false-admission rate

fx auto mode
A Vercel safety classifier Sharma says previously ran on gpt-5.6-luna. The September 16 post is the source for the 5 to 18 times figure.

TechCrunch covers Jev, and quotes a Vercel engineer who swapped Luna for it

Geometric mean
The four axis scores are multiplied and the fourth root is taken. A weak axis pulls the total down; Luna's cost score of 28.2 is why a 96.8 Intelligence row ranks 12th at 66.0.

JevBench scores typed decisions on four axes, and Jev's lead is a pricing story

GLiFormer
Knowledgator's encoder with task heads at inference time. jeff loads knowledgator/gliformer-large-v1.

LlamaIndex's Logan Markewich ships jeff, a local System One API on GLiFormer

head_max_len
Laya's option-prompt token budget. At default 256 on the multilingual checkpoint, 77 Banking77 labels get about 3 to 4 tokens each, which the README says is why accuracy falls to 0.425.

Laya, an Apache 2.0 System One stack, publishes where it beats Jev's numbers and where it does not

Header denylist
semgate.WithHeaderDenylist. Without it, Authorization and Cookie are sent to TypeSafe with the rest of the request.

semgate asks Jev about each HTTP request, then a live range to attack it

Helium
A browser reported to cause account registration failures during the initial onboarding period.

TypeSafe starts pulling people off the Jev waitlist

Hierarchical context
A proposed structure organizing execution history into a nested task tree to allow logarithmic search instead of sequentially reading complete logs.

Almeida's bet on Jevons paradox over chatbot price cuts

high_risk floor
In the AgentOS Jev strategy, a Noul at or above 0.7 forces the turn to tier c3 in code, so a short request to delete production data does not depend on the route Choice reading the words literally.

AgentOS posts a 121-turn routing table for an opt-in Jev strategy

hint mode
The jev-gateway path for Claude Code. A short suggestion is added to the request because a forced tool_choice would break extended thinking or the prompt cache. The model can ignore it.

jev-gateway asks Jev which tool a coding agent should call

Hit@1
Share of labelled queries whose correct document is ranked first. On McLaren's 41-query TypeSafe-docs set, jevsearch is 83% and the keyword pass alone is 41%.

jevsearch re-ranks keyword hits with Jev and scored 83% Hit@1 on the author's TypeSafe docs set

Hybrid override
Postgres plans first. If Jev's estimate is confident, that plan can be replaced. The thread reports this avoided the 10x regression from blind overrides.

A hybrid Jev overlay sped up Postgres joins 12% on JOB, and made the planner slower

Independent verifier
A check of the app's actual state (file locations and hashes, Calculator display) that the decision policy cannot read.

Jev macOS Loop clicks native Mac windows; Finder sorted nine files in 7.39 seconds

Injection Range
semgate-example, a Cloud Run service with six simulated injection sinks. With a TypeSafe key the guard answers 403; without a key every payload reaches the handler. No real side effects.

semgate asks Jev about each HTTP request, then a live range to attack it

input battery
The JevGuardrailAdvisor check on the user's request. A block returns before the chat model is called. The output battery checks the model's answer and does not retry.

Spring publishes a Java client that asks Jev before the chat model runs

Intelligence-per-cost
Almeida's term for the trade-off between computational cost and decision quality, prioritized over raw demonstration latency.

Almeida's bet on Jevons paradox over chatbot price cuts

Jaggedness
TypeSafe's word for uneven failure modes on jev-1.13. The docs page lists nine of them and says many will be fixed in later versions.

TypeSafe lists nine ways Jev 1.13 is still jagged, and asks for more

jeff
A self-hosted server that speaks TypeSafe's System One API and answers with Knowledgator's GLiFormer instead of Jev.

LlamaIndex's Logan Markewich ships jeff, a local System One API on GLiFormer

jev_ask
A Snowflake function that sends several named questions about one row in a single TypeSafe call, so the row is billed once.

Jevflake puts Jev in Snowflake: noul, choice, and score as SQL functions

jev_plan
A tool the agent calls with three to five hypotheses and a read-only test for each. Jev ranks those tests against a fresh namespace snapshot and does not run them.

SREGym put Jev on a Codex agent: 24 passes out of 50, up from 20

jev_stream
A DuckDB table function in duckdb-jev. It takes a subquery of row id, evidence, and questions, packs HTTP batches across chunks, and returns answers keyed by the supplied ids.

A native DuckDB extension batches Jev over SQL rows at about 2,300 rows per second on a live run

jev_submit
A review gate before diagnosis or mitigation is sent to the grader. Every required question had to reach probability 0.70 or the agent had to gather new evidence.

SREGym put Jev on a Codex agent: 24 passes out of 50, up from 20

jev-compaction
LiteLLM guardrail that asks Jev whether each older tool result still matters for the latest user question, then replaces results below relevance_threshold (default 0.2) with a short notice.

LiteLLM proxies Jev's evaluate endpoint and uses it to drop stale tool results

jev-latest
TypeSafe's alias for the current stable model release, currently routing to jev-1.13.0.

OpenRouter, Cloudflare, and Netlify start serving Jev

JevBench public subset
The 231 public tasks in JevBench, out of 534. Original 72, easy 48, hard 111. The other 303 tasks are private or judge-only, so a score on this slice is not the full benchmark composite.

Open-Jev's released 9B checkpoint answers 179 of 231 public JevBench tasks

JevBench Score
Benchmark Heaven's composite for typed-decision systems: Intelligence, Calibration, Speed, and Cost at 25% each, combined with a geometric mean. v1.2.2 reports Jev 1.13.0 at 75.3.

JevBench scores typed decisions on four axes, and Jev's lead is a pricing story

jevctl
npm package for Nasrallah-AL/jev-cli. The command is jev. It wraps TypeSafe's Jev as verify, screen, classify, extract, find, rerank, match, route, ask, compact, and batch.

jevctl puts Jev on the command line, with a compaction hook that keeps lines verbatim

jevel
A folder with one JEVEL.md. Frontmatter names the model, the required state fields, and the questions. jevelry reads that file. The body is for people.

jevelry turns a Jev call into a folder, a verdict, and a log

Jevflake
Luke Kranz's dbt package (and Terraform module) that creates Snowflake SQL functions for Jev's noul, choice, and score questions. Default model jev-1.13.0.

Jevflake puts Jev in Snowflake: noul, choice, and score as SQL functions

Jevis
A Dart package that runs on Flutter integration_test, sends the current screen and a registered action list to Jev, and executes the chosen widget action.

Jevis is a Flutter integration-test agent that only taps what you registered

JevJudge
A Spring AI TypeSafe builder. Each criterion is a Noul or a Score plus the threshold it must clear. Confidence under the default floor of 0.5 is reported as INCONCLUSIVE rather than FAILED.

Spring publishes a Java client that asks Jev before the chat model runs

Jevons paradox
The economic observation that increases in resource efficiency can expand total demand. The product is named around this economic hypothesis.

Almeida's bet on Jevons paradox over chatbot price cuts TypeSafe leaves stealth and ships Jev, a model that refuses to chat

jevsearch
Kyle McLaren's MIT shadcn registry block for site search. A lexical pass answers first; Jev re-ranks the top candidates in one request.

jevsearch re-ranks keyword hits with Jev and scored 83% Hit@1 on the author's TypeSafe docs set

Join order benchmark
JOB, a standard set of join-heavy queries over IMDB-style data, used here as the reported test. The thread does not list which query subset or Postgres version.

A hybrid Jev overlay sped up Postgres joins 12% on JOB, and made the planner slower

jsort
Khaled Eltokhy's MIT command, installed as jev-sort. It asks Jev which of two texts ranks higher on a plain-English comparative, fits a Bradley-Terry scale, and prints the file in that order with a standard error.

jsort ranks a file by asking Jev which of two texts is more of something

Judgment
One typed answer to one question about one document. 37 documents times 21 questions produced the 777 count.

Every ran 777 judgments in under a second and still wanted a better accuracy check

Judgments model
Jevflake's recommended dbt table: one stored answer per key and question, resent only when the row, the questions, or jevflake_model change.

Jevflake puts Jev in Snowflake: noul, choice, and score as SQL functions

jym
The command from syumai/jevyoumean. It wraps a CLI, and on an unknown subcommand asks Jev which documented command the token resembles, then can run that correction.

jym asks Jev which real subcommand a mistyped one was aiming at

keep-threshold
jev compact's default 0.5. A tool call or its result is kept when Jev's probability is at least this. LiteLLM's similar guardrail defaults to 0.2.

jevctl puts Jev on the command line, with a compaction hook that keeps lines verbatim

Kev
Jared Palmer's Apache 2.0 family of Jev-like decision models. Rank-16 LoRA plus a pointer head on Qwen bases. Serves POST /v1/systemone.

Jared Palmer's Kev family copies Jev's API on Qwen, and trails it on new sources

Late
A status indicating the model response missed the block deadline, resulting in a hold action without updating quotes.

A buy-or-sell every Monad block, with a mock default model

Laya
Convai Innovations' Apache 2.0 typed-decision engine. Three checkpoints plus a Router. JevBench v1.2.2 reports 70.1 for the Laya row against Jev 75.3.

Laya, an Apache 2.0 System One stack, publishes where it beats Jev's numbers and where it does not

Legal landing
A valid board placement calculated by the game engine. The model chooses exclusively among these options.

Tetris as a legal-move test: Jev picks from the landings you give it

LiteParse
The default local parser in DocJev. It extracts complete page text on the machine. Decision timing in the published pilots excludes this step.

DocJev classifies and splits real PDFs with Jev, 5.7x to 6.5x faster than Luna on the author's set

Literal reading
Answering the words in the instruction rather than the intent behind them. The docs say to put boundary cases in the criteria, or split the judgment into two questions.

TypeSafe lists nine ways Jev 1.13 is still jagged, and asks for more

llamajev
The MIT command in NakliTechie/llamacpp-jev. It launches or attaches to an unmodified llama-server and serves POST /v1/systemone by reading single-token label logprobs.

llamacpp-jev answers /v1/systemone from an unmodified llama-server, including images

LLM fallback
A small language model used exclusively when the chosen operation requires text input.

Browser Use wires Jev into the click loop: 7 seconds, $0.0039

LLM-as-a-judge
A generative model scores another system's trace. LangChain compared that pattern with Jev on five frozen weather-agent runs.

LangChain ran Jev as a judge: 500 matching pass/fail labels, and a $0.34 bill against Claude's $28

Local perception
ScreenCaptureKit, OmniParser CoreML, Apple Vision OCR, and accessibility labels run on the Mac. Pixels and coordinates are not sent to Jev.

Jev macOS Loop clicks native Mac windows; Finder sorted nine files in 7.39 seconds

LocalJev
GitHub Next's local HTTP server that accepts TypeSafe's System One request shape and answers with a model running behind an OpenAI-compatible Chat Completions API.

GitHub Next ships LocalJev, a prompted stand-in that speaks /v1/systemone

Macro accuracy
Accuracy with each of the 10 categories weighted equally. He reports 92.0% for Jev, 94.6% for Flash-Lite, and 96.9% for Flash.

Bryo ran 1,565 business emails: Jev at 96.4%, Gemini a point or two higher

MCP
Model Context Protocol. typesafe-mcp and jev-mcp expose Choice, Score, and Noul as tools callable by Claude Code or similar clients.

A week of Jev repos: routers, MCP, graphs, games

meriyah
A DeepSWE task whose unfiltered tool output was 1.46 million characters. Unfiltered 0/49 tests; filtered 49/49.

AstroHan ran 30 coding tasks with a Jev context filter: 25 passed, 22 without it

Mock model
The repository's default heuristic that simulates trading signals locally without requiring an API key.

A buy-or-sell every Monad block, with a mock default model

Model completion
A Noul on the current screen clearing goalThreshold. Reports mark this completionBasis as model unless a verify callback also passes, in which case it is model+assertion.

Jevis is a Flutter integration-test agent that only taps what you registered

ModelRouterMiddleware
An experimental middleware hook where Jev analyzes the latest user input to select which chat model processes the turn.

LangChain ships a Jev classifier, then parks it on the agent loop

n8n-nodes-jev-classification
MIT n8n community node from khmuhtadin. Classify, Score, Check, and Ask Questions against TypeSafe's Jev. Fixed classify mode emits one output per category plus Needs Review.

An n8n community node routes Jev answers onto category branches, including Needs Review

Native computer use
Clicking and dragging controls in a real macOS window rather than driving a browser DOM.

Jev macOS Loop clicks native Mac windows; Finder sorted nine files in 7.39 seconds

Needs Review
The extra classify output when confidence is below the node's threshold, default 0.5. Items still carry jev.needsReview when the node is told to send them to the best category anyway.

An n8n community node routes Jev answers onto category branches, including Needs Review

New sources
Palmer's out-of-domain split: datasets and policy rule types Kev was not trained on. The September 20 post puts Kev-8B at 79.6% there against Jev 85.7%.

Jared Palmer's Kev family copies Jev's API on Qwen, and trails it on new sources

Next-token logit read
Score the allowed answer tokens from one prefill, with no decode loop. SimpleJev reuses the shared prompt's KV cache across questions in the same request.

SimpleJev turns an open model into a classifier by reading the next-token logits

No state retention
Venice's claim that it keeps no copy of the state you send or the answers it returns, and that it does not train on that traffic.

Venice lists Jev on its API: same key, no copy of the state

Noul
TypeSafe's term for a boolean question, returning a probability score between 0 and 1 without a separate confidence field.

Hono's author routes HTTP by meaning, then warns you not to What Jev returns: System One in plain language

Open-Jev
Zefan Cai's Qwen3.5 LoRA adapters plus a trained decision head. The 2B and 9B checkpoints return Choice, Noul, and Score probabilities. A different project from SemIf, which previously used the name OpenJev.

Open-Jev's released 9B checkpoint answers 179 of 231 public JevBench tasks

opponent read
The Jev request in poker-table: a choice over range buckets, a noul for folding to each candidate bet size, and a noul for continue strength. The README says hole cards and chat are not in that request.

A Hold'em table asks Jev about the opponent and lets code pick the bet

Paid team share
Vercel's figure for the fraction of paying AI Gateway teams that called a model in a launch window. The Jev post uses nearly 13% at 24 hours.

Vercel says Jev is the fastest-adopted model in AI Gateway history, and makes it free until Sept 25

Parallel constrained decoding
Read the prompt once, then score the allowed answer tokens instead of generating a string. Nimble's Python code builds the typed result from those scores.

Bespoke Nimble is an open 9B Jev-style model: 90% on its own 324-example holdout

Parallel questions
Submitting multiple questions regarding the same state within one API request, distinct from classifying multiple different emails simultaneously.

One builder ran Jev on 1,500 of his own emails

Parallel sampling
Evaluating multiple declared questions over a single input state simultaneously without sequential token generation.

What Jev returns: System One in plain language

Pass rate
Successful attempts out of five per problem. The authors used this instead of time-to-diagnosis.

SREGym put Jev on a Codex agent: 24 passes out of 50, up from 20

Pass-through
LiteLLM's /typesafe/ prefix. The proxy forwards TypeSafe paths, including POST /v1/systemone, and returns TypeSafe's response unchanged.

LiteLLM proxies Jev's evaluate endpoint and uses it to drop stale tool results

pass@1
Whether the task passed on the single attempt that was run. 25/30 with the filter, 22/30 without, on the FrontierHarness 30.

AstroHan ran 30 coding tasks with a Jev context filter: 25 passed, 22 without it

Pilot Router
AgentOS's layer that picks a model tier for a turn, from c0 for a short reply to c3 for deep work. The Jev strategy is opt-in. The default remains the local MiniLM classifier.

AgentOS posts a 121-turn routing table for an opt-in Jev strategy

Planted defect
A problem inserted into synthetic text so the tester knows which passages should be flagged.

Every ran 777 judgments in under a second and still wanted a better accuracy check

Play credit
The hosted demo signs you in through Standard Agents and gives $0.25 of Jev spend. Local development uses your own TypeSafe key and skips login.

JevPilot is a playable driving sim that asks Jev which path to take

plpython3u
The untrusted Python language PostgreSQL needs for this extension. The README says the function performs external HTTPS work inside the statement, and that a rollback does not undo a request already sent.

pg-jev puts prompt_jev() in PostgreSQL and returns jsonb

Pointer head
Kev's scoring layer. It compares each option's </opt> hidden state with the question's <decide> hidden state, then softmaxes those scores.

Jared Palmer's Kev family copies Jev's API on Qwen, and trails it on new sources

port fidelity
In the laya-mlx README, whether the MLX port selects the same answer as upstream Laya on a fixed set of 63 validation questions. The README separates that check from accuracy against a human label or against Jev.

laya-mlx runs the Laya checkpoints on Apple Silicon

Post-only
An order that adds liquidity to the order book rather than filling immediately against existing orders, capturing the spread when matched.

A buy-or-sell every Monad block, with a mock default model

Precision-gated automation
Fit a confidence cutoff on validation for a target precision, then apply it to held-out traffic. Poly's 90% precision bar auto-handled 67.5% of items with Jev and 35.7% with ModernBERT.

Charly Poly put Jev next to open encoders, not GPT-5.6, and it lost on calibration

Promotional pricing
The Gateway model card's free input and output for typesafe-ai/jev, listed as ending September 25, 2026.

Vercel says Jev is the fastest-adopted model in AI Gateway history, and makes it free until Sept 25

prompt_jev
In mejiasd3v/pg-jev, a PostgreSQL function that sends text and typed questions to Jev and returns jsonb. The README's modes are a noul probability, a choice, a score, or several questions in one request.

pg-jev puts prompt_jev() in PostgreSQL and returns jsonb MotherDuck's prompt_jev() labels 100,000 AG News rows in 40 seconds for $0.50

Prompted probability
A probability vector the generative model writes as JSON, then LocalJev validates and normalizes. The README says this is not mathematically equivalent to reading logits.

GitHub Next ships LocalJev, a prompted stand-in that speaks /v1/systemone

Quality-score variance
How much a judge's continuous 0–1 quality score moved when the agent output was unchanged. LangChain reports Jev at 0.0000149 mean per-case variance.

LangChain ran Jev as a judge: 500 matching pass/fail labels, and a $0.34 bill against Claude's $28

Question isolation
Running a separate encoder pass per noul so other questions in the same request cannot change that probability. jeff defaults to isolating nouls.

LlamaIndex's Logan Markewich ships jeff, a local System One API on GLiFormer

raw label softmax
llamajev's probabilities are a softmax over the logprobs of the option labels. The README says they are not TypeSafe's calibrated RLCD probabilities.

llamacpp-jev answers /v1/systemone from an unmodified llama-server, including images

Reference-label accuracy
Share of emails where the model's top category matches the label Mudholkar assigned. The 96.4% figure is this metric on 1,565 items, including 364 synthetic mails.

Bryo ran 1,565 business emails: Jev at 96.4%, Gemini a point or two higher

Reference-label agreement
Share of examples where the model's top answer matches the synthetic label. The 324-example table is this metric, not calibration against human judges.

Bespoke Nimble is an open 9B Jev-style model: 90% on its own 324-example holdout

Reject filter
Skip the slower model when Jev's P(yes) sits under a line set from known positives. On Kumar's page gates the line came from half the lowest P(yes) among real positives, checked by rotating production runs.

Aman Kumar ran about 16,000 Jev calls and called it a filter, not a classifier

RLCD
Reinforcement Learning for Calibrated Decisions, TypeSafe's training methodology intended to align model confidence scores with true outcome probabilities.

TypeSafe leaves stealth and ships Jev, a model that refuses to chat

Router
Laya's entry point that detects script and language, then dispatches to english, multilingual, or typed-decisions. Lazy load without preload can cost 7 to 10 s on a language switch.

Laya, an Apache 2.0 System One stack, publishes where it beats Jev's numbers and where it does not A week of Jev repos: routers, MCP, graphs, games

Runnable
LangChain's standard unit of composition that can be executed, batched, or streamed across agent pipelines.

LangChain ships a Jev classifier, then parks it on the agent loop

Schema
The structured Choice definition representing possible piece landings, preventing the model from outputting invalid moves.

Tetris as a legal-move test: Jev picks from the landings you give it

Score
A numeric evaluation based on a custom rubric, returning an estimated position, distribution, and confidence level.

What Jev returns: System One in plain language

Selectivity
The fraction of rows a filter is expected to keep. Malis asked Jev for this number and let Postgres plan from it.

A hybrid Jev overlay sped up Postgres joins 12% on JOB, and made the planner slower

Semantic route
An HTTP route handler matched by a descriptive phrase (such as 'a request from an AI agent') instead of an exact method and URL path.

Hono's author routes HTTP by meaning, then warns you not to

semgate
mizutani's Go net/http middleware. It asks Jev typed questions about an incoming request and hands the answers to your function, which may block, route, or call next.

semgate asks Jev about each HTTP request, then a live range to attack it

SemIf
Theodore Lee's open project, formerly OpenJev, that reads typed option probabilities from a local model. JevBench v1.2.2 reports 74.6 for the Qwen3.5-4B row.

SemIf, a 4B open rebuild that reads option logits, sits 0.7 behind Jev on JevBench

Set A
Per-headline questions: a desk, a story type, and six 0-4 scores (magnitude, velocity, novelty, window, heat, risk) that feed a newsworthiness mark out of 10.

Jev scored 384 morning headlines in 25 seconds; Opus 5 finished four

Set B
Per-company questions on standing and journalist shape, then a tier. A story can land on several of the 15 illustrative desks, or none.

Jev scored 384 morning headlines in 25 seconds; Opus 5 finished four

settle
After each action, dejevu waits until the DOM is quiet and the requests that action started have finished, with a cap per action type. The README contrasts that with fixed waits, and says jev-ultrafast discards 4 to 6 of 17 calls because the page changed under them.

dejevu books the same Flights search faster than the published Jev demo, on another machine

Shadow mode
Classify and log the label, then act as if control_plane is off. The PR default for live.classify.shadow is true.

Sam Saffron tries Jev as a hidden control plane for GPT Live in term-llm

Sieve
A filter placed before a larger model to discard unnecessary context according to calibrated criteria.

Compaction without summarization: score the tool calls and drop them

signup pause
TypeSafe's September 22 statement that new Jev accounts are temporarily not being created, while accounts that already signed up keep working. The post does not give a reopen time.

TypeSafe pauses new Jev signups

SimpleJev
Featherless's Apache 2.0 classifier server. It reads next-token logits for allowed labels and returns Choice, Score, or Noul JSON. JevBench v1.2.8 lists Qwen3.8-27B at 67.3.

SimpleJev turns an open model into a classifier by reading the next-token logits

Skill file
A SKILL.md instruction file read by coding agents. TypeSafe's file prompts agents to fetch updated documentation rather than relying on stale parameters.

Hono's author routes HTTP by meaning, then warns you not to

spot bank
46 handwritten Hold'em spots in the poker-table repo. The Jev mode repeats them three times through OpenRouter on typesafe/jev-1.13. The README gives a median near one second and does not print an accuracy percentage.

A Hold'em table asks Jev about the opponent and lets code pick the bet

SREGym-Lite
A smaller SREGym problem set meant to run on modest clusters. This experiment used 10 of those problems, not the full 21-problem lite suite.

SREGym put Jev on a Codex agent: 24 passes out of 50, up from 20

Stagehand
Browserbase's library for automating remote browser sessions through structured interaction steps.

Stagehand, Chrome, and a11y trees: computer use as a Jev question

Starting credit
The $5 TypeSafe posts for every new account, described as about 120 million tokens at $0.042 per million input.

TypeSafe drops the Jev waitlist; new accounts start with $5

State
The input data provided to Jev, such as a single email or batch. The calling application retrieves and supplies the data.

One builder ran Jev on 1,500 of his own emails

status.typesafe.ai
TypeSafe's public status page for api.typesafe.ai and console.typesafe.ai. The September 21 incident is titled Console is unavailable.

TypeSafe posted an API and console outage, then said it was back

Structural invariant
A relationship you might expect between two questions, such as P(yes) on a Choice matching a Noul, or a Noul and its negation summing to 1. The docs say jev-1.13 does not guarantee those identities.

TypeSafe lists nine ways Jev 1.13 is still jagged, and asks for more

Structured read
OpenJev's path, as LocalJev describes it: a one-step DiffusionGemma pass that needs seeded canvases, read-only denoising, and selected-token logprobs, which stock oMLX does not expose.

djev reads DiffusionGemma as a decision API, and sits third on JevBench GitHub Next ships LocalJev, a prompted stand-in that speaks /v1/systemone

System One
TypeSafe's term for a model returning structured decisions with probabilities rather than text output, inspired by Daniel Kahneman's distinction between fast and slow cognition.

TypeSafe leaves stealth and ships Jev, a model that refuses to chat

Target element
The specific page element receiving the selected action. The repository states that both heads share a request to avoid mismatched actions and targets.

Browser Use wires Jev into the click loop: 7 seconds, $0.0039

teacher average
The blog's name for the reference on typed-decisions: the mean of three samples from teacher models. It says two single samples agree 73.5% of the time, so a model can match the average more often than one teacher matches another.

meraGPT posts Decider 1 ahead of Jev on a teacher-labelled set

Threshold
The minimum score required for a match (defaulting to 0.5). The first description whose Noul probability exceeds this value triggers its handler.

Hono's author routes HTTP by meaning, then warns you not to

TypeSafeClassifier
A LangChain Runnable that submits state alongside defined Choice, Score, or Noul queries to Jev, returning structured results rather than text generation.

LangChain ships a Jev classifier, then parks it on the agent loop

Verbatim keep
Retaining original tool outputs unchanged in context without summarization, preventing drift in variable names or error logs.

Compaction without summarization: score the tool calls and drop them

verdict
act, mark, or fall_back, chosen from thresholds in the jevel. act means the probability clears the bar to use the answer. mark means use it and flag it. fall_back means do what the code did before the call.

jevelry turns a Jev call into a folder, a verdict, and a log

verified run
In dejevu, a run counts only when code checks the final page. The README says the model announcing that it is done is not the check. Flights must show results. Wikipedia must land on the exact article URL.

dejevu books the same Flights search faster than the published Jev demo, on another machine

Waitlist
TypeSafe's early-access queue from launch through September 20, 2026, when the company posted that Jev was open with no waitlist.

TypeSafe drops the Jev waitlist; new accounts start with $5 TypeSafe starts pulling people off the Jev waitlist

web context
In the Exa demo, search snippets placed into the state Jev judges. The page says Jev is asked the same yes/no question with that state and without it. No scored pair was printed on the page we loaded.

Exa posts a demo that puts web results into Jev's state

Whole-document read
A question whose answer needs the page as a unit, not a short snippet. Kumar's 135-row extraction gate is the example: Jev matched later outcomes 68.6% and the probability swung by 0.5 between runs on the same page.

Aman Kumar ran about 16,000 Jev calls and called it a filter, not a classifier

Workers AI
Cloudflare's serverless inference runtime, executing queries via env.AI.run('typesafe/jev', { state, questions }).

OpenRouter, Cloudflare, and Netlify start serving Jev

x402 wallet
A prepaid USDC path Venice lists next to a Bearer API key for authenticating decision calls.

Venice lists Jev on its API: same key, no copy of the state

Zero-probability miss
The model assigns probability 0 to the correct label. Poly reports this on 6.6% of Jev's decisions and on none of 1,155 decisions from each of the three open encoders. A zero cannot be recovered with a threshold.

Charly Poly put Jev next to open encoders, not GPT-5.6, and it lost on calibration