{"slug": "is-jev-that-good", "title": "Is Jev That Good?", "summary": "Diogo Almeida, who says he co-invented ChatGPT, released a new frontier AI model called Jev through his company typesafe.ai, claiming it is 20-200x faster and 40-400x cheaper than existing models with output tokens free forever at $0.042 per million input tokens. Almeida said he spent the last 2 years in stealth developing a new training method, RLCD, and that Jev is a \"frontier composable intelligence optimized for decisions\" that cannot generate text. The release blog is at typesafe.ai/blog and a waitlist is open for early access.", "body_md": "Diogo Almeida on X: \"After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?\nI’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev\n• 20-200x faster\n• 40-400x … / X\n\nDiogo Almeida on X: \"After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?\nI’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev\n• 20-200x faster\n• 40-400x cheaper (w/ output tokens free)\n• Frontier composable intelligence optimized for decisions\nAFAICT the shortest path to AI-based economic revolution\"\n\nAfter co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?\nI’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev\n• 20-200x faster\n• 40-400x cheaper (w/ output tokens free)\n• Frontier composable intelligence optimized for decisions\nAFAICT the shortest path to AI-based economic revolution\n\nAfter co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?\nI’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev\n• 20-200x faster\n• 40-400x cheaper (w/ output tokens free)\n• Frontier composable intelligence optimized for decisions\nAFAICT the shortest path to AI-based economic revolution\n\nThe gains aren’t free: Jev can't generate text\nComparing Jev vs LLMs side-by-side makes the trade-off clear\nFun fact: replacing sequential computation with parallel is the same way Transformers leapfrogged RNNs\n\nWe believe that the future is code + AI, so made workflow evals to reflect that\nJev costs: $42 / BILLION input tokens ($0.042 / MTok) and output tokens are free (forever - they’re too cheap to meter with our new architecture)\nJev is named after Jevons paradox and off theShow more\n\nGame: race from one Wikipedia page to another using only links\nChallenge: choosing between hundreds to thousands of links\nShows not just intelligence-per-second, but also the compounding benefits of not hallucinating with high-cardinality choices\n\nExtraordinary claims require extraordinary evidence so check out our release blog for more technical info: typesafe.ai/blog/introduci…\nJoin our waitlist for early access: typesafe.ai\nHave technical chats and meme with us on Discord (rumors are good memers skip theShow more", "url": "https://wpnews.pro/news/is-jev-that-good", "canonical_source": "https://twitter.com/completeskeptic/status/2099925682726002904", "published_at": "2026-09-16 05:38:03+00:00", "updated_at": "2026-09-16 06:07:45.867566+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-research", "ai-startups", "large-language-models", "ai-products"], "entities": ["Diogo Almeida", "Jev", "ChatGPT", "typesafe.ai", "RLCD", "AGI", "X"], "alternates": {"html": "https://wpnews.pro/news/is-jev-that-good", "markdown": "https://wpnews.pro/news/is-jev-that-good.md", "text": "https://wpnews.pro/news/is-jev-that-good.txt", "jsonld": "https://wpnews.pro/news/is-jev-that-good.jsonld"}}