--- title: Jev: The New AI Model That's Breaking The Internet (Full... description: A new AI model called Jev (referred to as 'Jeff' throughout) is presented as a radically faster, cheaper alternative to LLMs for decision-making tasks. Rather... canonical: https://daily.dev/posts/jev-the-new-ai-model-that-s-breaking-the-internet-full-tutorial--nyqdaotb2 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Jev: The New AI Model That's Breaking The Internet (Full Tutorial) | daily.dev og:description: A new AI model called Jev (referred to as 'Jeff' throughout) is presented as a radically faster, cheaper alternative to LLMs for decision-making tasks. Rather... og:url: https://daily.dev/posts/jev-the-new-ai-model-that-s-breaking-the-internet-full-tutorial--nyqdaotb2 og:image: https://api.daily.dev/og/posts/NyqDaoTB2.png og:image:alt: Jev: The New AI Model That's Breaking The Internet (Full Tutorial) og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Jev: The New AI Model That's Breaking The Internet (Full Tutorial) **[YouTube](https://daily.dev/sources/youtube)** · 22 min read · 0 upvotes · 0 comments ## Summary A new AI model called Jev (referred to as 'Jeff' throughout) is presented as a radically faster, cheaper alternative to LLMs for decision-making tasks. Rather than generating text token by token, Jev outputs classification probabilities, making it 20-200x faster and 40-400x cheaper than frontier LLMs, though it cannot write text, explain itself, or reason step by step. The creator demonstrates obtaining API access via typesafe.ai (currently waitlisted) and builds three example projects using Cursor: a voice-controlled web browser that executes commands in real time, an improved personal memory/knowledge-retrieval system that cuts token usage by up to 98%, and a YouTube title/thumbnail scoring and ranking tool that predicts video performance. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the difference between an LLM and a classification model like Jev for making decisions in software? An LLM generates a text answer token by token, which is slow (seconds) and costly, while a classification-based model like Jev outputs a probability score against predefined categories in milliseconds. Jev cannot write text, explain itself, or reason step by step, but it can classify, score, rank, or route decisions at roughly 100x the speed and a fraction of the cost, making it suited for the many if-then decisions inside software rather than open-ended generation. _Developers weighing when to use fast classifiers versus LLMs can track approaches like this via daily.dev._ ### How can a classification model reduce token usage when retrieving memory from a large set of markdown files? Instead of an LLM guessing which file holds an answer and reading the whole file, a classification model can answer a batch of small yes/no or pick-one questions about files, sections, and bullets in one request, returning probabilities in milliseconds. In one demonstrated system this cut a memory-recall request from about 13,000 tokens to roughly 2,756 tokens (about 80% fewer), and in another case down to 293 tokens (98% fewer), with some queries needing zero tokens when the answer wasn't present at all. _Engineers building efficient retrieval systems can follow token-saving techniques like this on daily.dev._ --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents) [View this post on daily.dev](https://daily.dev/posts/jev-the-new-ai-model-that-s-breaking-the-internet-full-tutorial--nyqdaotb2) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Jev: The New AI Model That's Breaking The Internet (Full Tutorial)","url":"https://daily.dev/posts/jev-the-new-ai-model-that-s-breaking-the-internet-full-tutorial--nyqdaotb2","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/jev-the-new-ai-model-that-s-breaking-the-internet-full-tutorial--nyqdaotb2"},"datePublished":"2026-09-21T02:32:52.906Z","dateModified":"2026-09-21T17:29:14.823Z","description":"A new AI model called Jev (referred to as 'Jeff' throughout) is presented as a radically faster, cheaper alternative to LLMs for decision-making tasks. Rather...","image":"https://i.ytimg.com/vi/Nq_lu5QT-fI/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/Nq_lu5QT-fI/sddefault.jpg","isAccessibleForFree":true,"articleSection":"YouTube","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"YouTube","logo":"https://media.daily.dev/image/upload/s--W5zMumpP--/f_auto/v1725350181/logos/youtube","url":"https://daily.dev/sources/youtube"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/jev-the-new-ai-model-that-s-breaking-the-internet-full-tutorial--nyqdaotb2","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ai-agents","timeRequired":"PT22M","video":{"@type":"VideoObject","name":"Jev: The New AI Model That's Breaking The Internet (Full Tutorial)","description":"A new AI model called Jev (referred to as 'Jeff' throughout) is presented as a radically faster, cheaper alternative to LLMs for decision-making tasks. Rather...","thumbnailUrl":"https://i.ytimg.com/vi/Nq_lu5QT-fI/sddefault.jpg","uploadDate":"2026-09-21T02:32:52.906Z","duration":"PT22M","url":"https://api.daily.dev/r/NyqDaoTB2","embedUrl":"https://www.youtube.com/embed/Nq_lu5QT-fI"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"YouTube","item":"https://daily.dev/sources/youtube"},{"@type":"ListItem","position":3,"name":"Jev: The New AI Model That's Breaking The Internet (Full Tutorial)"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/jev-the-new-ai-model-that-s-breaking-the-internet-full-tutorial--nyqdaotb2#faq","mainEntity":[{"@type":"Question","name":"What is the difference between an LLM and a classification model like Jev for making decisions in software?","acceptedAnswer":{"@type":"Answer","text":"An LLM generates a text answer token by token, which is slow (seconds) and costly, while a classification-based model like Jev outputs a probability score against predefined categories in milliseconds. Jev cannot write text, explain itself, or reason step by step, but it can classify, score, rank, or route decisions at roughly 100x the speed and a fraction of the cost, making it suited for the many if-then decisions inside software rather than open-ended generation. Developers weighing when to use fast classifiers versus LLMs can track approaches like this via daily.dev."}},{"@type":"Question","name":"How can a classification model reduce token usage when retrieving memory from a large set of markdown files?","acceptedAnswer":{"@type":"Answer","text":"Instead of an LLM guessing which file holds an answer and reading the whole file, a classification model can answer a batch of small yes/no or pick-one questions about files, sections, and bullets in one request, returning probabilities in milliseconds. In one demonstrated system this cut a memory-recall request from about 13,000 tokens to roughly 2,756 tokens (about 80% fewer), and in another case down to 293 tokens (98% fewer), with some queries needing zero tokens when the answer wasn't present at all. Engineers building efficient retrieval systems can follow token-saving techniques like this on daily.dev."}}]} ```