--- title: Why Featherless says you don’t need a tank to deliver a... description: Featherless has launched Simple Jev, an open-source library that converts open-source AI models into fast zero-shot classification engines, letting apps return... canonical: https://daily.dev/posts/why-featherless-says-you-don-t-need-a-tank-to-deliver-a-pizza-zwinbarlj twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Why Featherless says you don’t need a tank to deliver a pizza | daily.dev og:description: Featherless has launched Simple Jev, an open-source library that converts open-source AI models into fast zero-shot classification engines, letting apps return... og:url: https://daily.dev/posts/why-featherless-says-you-don-t-need-a-tank-to-deliver-a-pizza-zwinbarlj og:image: https://api.daily.dev/og/posts/zWInbArlj.png og:image:alt: Why Featherless says you don’t need a tank to deliver a pizza 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. # Why Featherless says you don’t need a tank to deliver a pizza **[The New Stack](https://daily.dev/sources/newstack)** · 6 min read · 0 upvotes · 0 comments ## Summary Featherless has launched Simple Jev, an open-source library that converts open-source AI models into fast zero-shot classification engines, letting apps return categorical or binary decisions without generating conversational text. Built on top of TypeSafe's closed-source Jev launched earlier the same month, Simple Jev adds support for open models and, on Featherless's hosted endpoints, vision inputs via Gemma or Qwen. Pricing starts at $0.03 per million input tokens with free output tokens, undercutting Jev's $0.042 per million input rate, though Featherless's own Qwen-based classifiers list at $0.28-$0.30. CEO Eugene Cheah argues frontier LLMs are overkill for simple classification tasks, comparing it to using a tank to deliver a pizza, and positions Simple Jev as reviving pre-ChatGPT classifier techniques with a modern API. Free public demo endpoints are capped at 2,000 tokens of context and two requests per second, and the library is available on GitHub. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is Featherless's Simple Jev and how does pricing compare to TypeSafe's Jev? Simple Jev is an open-source library from Featherless that turns open-source AI models into zero-shot classification engines, stopping generation before text output and returning probability scores for allowed categories instead. It costs $0.03 per million input tokens with free output tokens, versus $0.042 per million input tokens for TypeSafe's closed-source Jev, though this is a beta floor price that may rise. _Teams weighing lightweight classifiers against frontier LLM pricing can track releases like this one on daily.dev._ ### How does zero-shot classification with Simple Jev differ from using a large language model to classify text? Simple Jev never generates a written answer; it halts the model at the decision point, reads the logit scores for each allowed option, and outputs probabilities directly, which cuts latency and compute cost compared to having an enormous frontier LLM write out a classification response. It currently supports vision-based classification only with Gemma or Qwen models on Featherless's hosted endpoints. _Developers choosing between heavyweight LLMs and dedicated classifiers can follow tooling comparisons like this on daily.dev._ ### How much does classifying data with Featherless's Simple Jev cost per decision? A typical classification decision using 500 to 1,200 tokens costs roughly three thousandths of a cent at the $0.03 per million input token rate, or about $15-35 per million decisions, with output tokens free. Featherless's own Qwen-based classifiers are listed higher, at $0.28 and $0.30 per million input tokens. _Engineers estimating inference costs for classification workloads can find pricing breakdowns like this on daily.dev._ --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#ai-inference](https://daily.dev/tags/ai-inference) [View this post on daily.dev](https://daily.dev/posts/why-featherless-says-you-don-t-need-a-tank-to-deliver-a-pizza-zwinbarlj) ```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":"Why Featherless says you don’t need a tank to deliver a pizza","url":"https://daily.dev/posts/why-featherless-says-you-don-t-need-a-tank-to-deliver-a-pizza-zwinbarlj","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/why-featherless-says-you-don-t-need-a-tank-to-deliver-a-pizza-zwinbarlj"},"datePublished":"2026-09-29T19:40:37.764Z","dateModified":"2026-09-29T19:41:46.093Z","description":"Featherless has launched Simple Jev, an open-source library that converts open-source AI models into fast zero-shot classification engines, letting apps return...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/587619a204992f5c4a5df6d6515064fb?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/587619a204992f5c4a5df6d6515064fb?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"The New Stack","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":"The New Stack","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/newstack","url":"https://daily.dev/sources/newstack"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/why-featherless-says-you-don-t-need-a-tank-to-deliver-a-pizza-zwinbarlj","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,ai-inference","timeRequired":"PT6M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"The New Stack","item":"https://daily.dev/sources/newstack"},{"@type":"ListItem","position":3,"name":"Why Featherless says you don’t need a tank to deliver a pizza"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/why-featherless-says-you-don-t-need-a-tank-to-deliver-a-pizza-zwinbarlj#faq","mainEntity":[{"@type":"Question","name":"What is Featherless's Simple Jev and how does pricing compare to TypeSafe's Jev?","acceptedAnswer":{"@type":"Answer","text":"Simple Jev is an open-source library from Featherless that turns open-source AI models into zero-shot classification engines, stopping generation before text output and returning probability scores for allowed categories instead. It costs $0.03 per million input tokens with free output tokens, versus $0.042 per million input tokens for TypeSafe's closed-source Jev, though this is a beta floor price that may rise. Teams weighing lightweight classifiers against frontier LLM pricing can track releases like this one on daily.dev."}},{"@type":"Question","name":"How does zero-shot classification with Simple Jev differ from using a large language model to classify text?","acceptedAnswer":{"@type":"Answer","text":"Simple Jev never generates a written answer; it halts the model at the decision point, reads the logit scores for each allowed option, and outputs probabilities directly, which cuts latency and compute cost compared to having an enormous frontier LLM write out a classification response. It currently supports vision-based classification only with Gemma or Qwen models on Featherless's hosted endpoints. Developers choosing between heavyweight LLMs and dedicated classifiers can follow tooling comparisons like this on daily.dev."}},{"@type":"Question","name":"How much does classifying data with Featherless's Simple Jev cost per decision?","acceptedAnswer":{"@type":"Answer","text":"A typical classification decision using 500 to 1,200 tokens costs roughly three thousandths of a cent at the $0.03 per million input token rate, or about $15-35 per million decisions, with output tokens free. Featherless's own Qwen-based classifiers are listed higher, at $0.28 and $0.30 per million input tokens. Engineers estimating inference costs for classification workloads can find pricing breakdowns like this on daily.dev."}}]} ```