--- title: "Startup TypeSafe AI Launches Low-Hallucination \"Jev\" Model" type: "News" locale: "en" url: "https://longbridge.com/en/news/300109181.md" description: "TypeSafe AI has launched \"Jev,\" a low-hallucination model positioned as a cost-effective and more efficient alternative to products from OpenAI and Anthropic, primarily targeting software developers. The company, currently valued at $200 million, has attracted billions of dollars in financing offers due to the model's impact, with a potential valuation reaching $10 billion. This move aims to reduce AI deployment costs and may exert pressure on the business models of mainstream large language models" datetime: "2026-09-25T07:23:07.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/300109181.md) - [en](https://longbridge.com/en/news/300109181.md) - [zh-HK](https://longbridge.com/zh-HK/news/300109181.md) generator: "portal-rs" --- # Startup TypeSafe AI Launches Low-Hallucination "Jev" Model A startup recently valued at $200 million is attracting financing offers worth billions of dollars after launching an AI tool claimed to be more efficient than products from OpenAI and Anthropic. TypeSafe AI was founded by Diogo Almeida, a former OpenAI researcher. The company just released its AI model, named "Jev," last week, positioning it as a cheaper alternative for performing certain tasks rather than relying on traditional large language models (LLMs) like ChatGPT and Claude. Jev is almost entirely oriented toward software developers. Since the company ended its so-called "stealth mode" and made its public debut, it has rapidly garnered significant attention and gone viral on social media. Its product introduction video posted on social platforms received 40 million views in less than a week. According to insiders, potential investors have presented financing proposals to TypeSafe AI, with valuations potentially reaching $10 billion or even higher. This enthusiasm reflects a growing concern in the corporate world: the cost of deploying AI is rising rapidly. TypeSafe is betting on the idea that many daily tasks currently handled by expensive general-purpose models can actually be completed by cheaper, more specialized systems. If this concept gains market acceptance, it could put pressure on the business models of leading AI companies such as OpenAI and Anthropic. **Does AI Need to Be More Specialized?** Almeida stated that he conceived the idea for Jev four years ago while working at OpenAI. At the time, he began to wonder: Are chatbots really suitable for all the tasks AI might be asked to perform? He recalled: "Assuming an AI-driven economic revolution truly occurs, how many of all AI requests are for human consumption... and how many are actually for computer consumption?" TypeSafe's core argument is: Large language models were initially designed for communication with humans but are not well-suited for the programmatic, repetitive, and high-frequency tasks that global enterprises are trying to automate through AI. Almeida stated: "ChatGPT is much smarter than I am. But strangely, the jobs with huge economic incentives that are most worthy of automation are not currently being automated. Because AI currently performs poorly in these areas." **How Does Jev Work?** Jev does not generate sentences, explain text, or interpret images. Instead, its goal is to make decisions quickly and at low cost within software applications. The "classification" tasks it can perform include: determining whether a request should be approved, rejected, or sent for manual review; automatically assigning customer service tickets; conducting insurance risk assessments; and evaluating credit risk. The model is primarily aimed at developers and is designed to be embedded in software backends rather than providing a consumer-facing chat interface. One developer integrated Jev into a joke website named "AskJev," paying homage to the now-defunct search engine Ask Jeeves, which subsequently went viral on social media. **Cost Advantage** James Hardiman, General Partner at Silicon Valley venture capital firm DCVC, stated that DCVC led TypeSafe's recent funding round. He said that because Jev significantly reduces computing costs, the company is already profitable, with model usage costs "orders of magnitude lower" than leading large language models. Hardiman pointed out that an AI tool like Jev, designed to handle very common tasks, entered the market just as public concern over the potential risks of advanced AI was increasing. He said: "Just a week ago, everyone was discussing AI doomsday scenarios... To some extent, Jev has made this discussion more rational." Reducing the cost of AI responses is key to Jev's market appeal. The name Jev is derived from the "Jevons paradox," a concept in economics: When a resource becomes cheaper or more efficient, total consumption may actually increase because new use cases continually emerge. Differences from Large Language Models Large language models typically require long chains of reasoning, which consume substantial computing resources. TypeSafe states that Jev adopts a different approach: It quickly selects results from a limited range of answers based on probability calculations. Its underlying method is closer to early machine learning systems. Such systems are typically more deterministic, producing fixed outputs for given inputs, thus avoiding the "hallucination" problems common in large language models. Due to lower computational requirements, Jev consumes fewer "tokens" (the computational unit for model text processing), thereby reducing costs. TypeSafe claims its cost per query is about one-hundredth that of large language models, and its processing speed is faster. The company's pricing is approximately: Jev: ~4.2 cents per million tokens; Large Language Models: potentially several dollars per million tokens. **Market Response** AI development platform Vercel stated that within 24 hours of Jev's launch, interest generated from its paid developer accounts exceeded that of any previous model release, including leading models from OpenAI and Anthropic. Another AI model aggregation platform, OpenRouter, reported that the number of tokens processed by Jev more than tripled over the weekend. Andrej Karpathy, OpenAI co-founder who recently joined Anthropic, stated on social media that Jev seems to have captured the market's latent demand for "simple, cheap, fast decision-making models." He said that leading AI companies have long focused on pursuing higher levels of intelligence, leaving this area "underinvested." Skepticism: Is Jev Truly Revolutionary? However, TypeSafe's secrecy regarding Jev's training methods, along with its similarities to existing technologies, has led some industry insiders to question the extent of its innovation. Anastasios Angelopoulos, Co-founder and CEO of AI model evaluation platform Arena, stated: "I am unclear on how these models differ from traditional 'zero-shot classifiers,' which are actually quite mature technology." Currently, Meta, Google, and Hugging Face all offer AI tools capable of classifying data they were not explicitly trained on. TypeSafe has not disclosed the specific training methods for Jev. The company stated that the model was trained using open-weight models and computer-generated "synthetic data," but relevant details remain strictly confidential. **Funding and Future Plans** Almeida stated that TypeSafe completed a $40 million seed funding round more than a year ago, at which time the company was valued at $200 million. He declined to disclose details of the new funding round or specific investors but confirmed: "Investors are knocking on our door like crazy." Almeida stated: "We are a group of missionaries, a spark of revolution... Therefore, as we scale the company, how to maintain our original intent and not lose our way is something we must seriously consider." ### Related Stocks - [ANTH.NA](https://longbridge.com/en/quote/ANTH.NA.md) - [OpenAI.NA](https://longbridge.com/en/quote/OpenAI.NA.md) - [META.US](https://longbridge.com/en/quote/META.US.md) - [GOOGL.US](https://longbridge.com/en/quote/GOOGL.US.md) - [GOOG.US](https://longbridge.com/en/quote/GOOG.US.md) ## Related News & Research - [OpenAI and Anthropic discover AI safety incidents on a scale far beyond what they’ve disclosed](https://longbridge.com/en/news/300209428.md) - [Anthropic's $518 billion AI buildout hinges largely on deals that cannot be canceled, filing shows](https://longbridge.com/en/news/300415240.md) - [Anthropic rolls out second Claude 5.5 model as it builds toward IPO](https://longbridge.com/en/news/300336765.md) - [EXCLUSIVE-Anthropic leaders to control AI lab via 'Founder LLC' to promote public good over market forces](https://longbridge.com/en/news/300358544.md) - [Another OpenAI Sandbox Fails as AI Agent Gains Internet Access](https://longbridge.com/en/news/300196925.md) --- > **Disclaimer: This article is for reference only and does not constitute any investment advice.**