ZeroHour
DashboardNewsCVEsExploitsIncidentsAIBriefAnalyticsEntitiesWatchlistIOCsAsk the feedSourcesFeeds & APIKiosk mode
Organization

TypeSafe AI

3 mentions in 7 days · 3 in 30 days · 3 total · first seen · last

Search everything

Timeline

TypeSafe AI Releases Jev: A System One Model That Returns Typed, Calibrated Decisions Instead of Text

TypeSafe AI launched Jev, a transformer-based 'System One' model API returning typed, calibrated decisions with probabilities, claiming 193x faster and 445x cheaper than LLMs.

TypeSafe AI released Jev, a transformer-based 'System One Model' that returns typed decisions with probabilities and confidence values instead of generated text. The API exposes three question primitives — Choice (up to 255 options), Score, and Noul (binary probability) — processed in parallel against a shared state and trained via Reinforcement Learning for Calibrated Decisions (RLCD). Pricing is $42 per billion input tokens with free output; vendor-run benchmarks claim 193.6x faster and 444.6x cheaper responses than GPT-5.6 Terra, though reference answers come from averaging other models and gains are unverified. Jev is in hosted early access behind a waitlist with no published weights, parameter count, or self-hosting option; community projects include agent guardrails, browser and phone agents, and Postgres natural-language filters.

MarkTechPost · 7h agoModel release 2 sources

Laya the open source version of Jev

Convai Innovations released Laya, an open-source Apache 2.0 family of non-autoregressive decision models claiming 6-8x faster inference than TypeSafe AI's Jev.

Convai Innovations launched Laya, a family of non-autoregressive 'System 1' decision models that output calibrated probabilities over schemas instead of generating text. It ships three Apache 2.0 checkpoints on Hugging Face: laya (421M, ModernBERT-large), laya-multilingual (322M, mmBERT-base, 100+ languages), and laya-typed-decisions (421M, 0.766 accuracy on security alerts), running in 32.8 ms on a single GPU (7.2 ms batched). The release positions Laya as an open alternative to TypeSafe AI's Jev, a commercial decision API priced at $0.042 per million input tokens with ~150 ms responses. The author cites prior arXiv papers (2503.23303, 2510.01237) and a 51-language MASSIVE benchmark sweep showing English-only encoders fail on non-Latin scripts (Khmer: 0.000 accuracy at 0.952 confidence).

Hacker News · security · 15h agoModel release 4 sourcesHN 22↑ · 3 comments

A new kind of AI model from a ChatGPT inventor is thrilling developers

TypeSafe AI, founded by ex-OpenAI RLHF inventor Diogo Almeida, released Jev, a non-LLM transformer outputting calibrated probabilities for cheap software automation.

TypeSafe AI, a startup founded two years ago by former OpenAI researcher Diogo Almeida, who helped invent RLHF, released a transformer-based model called Jev that is not a large language model. Instead of text, Jev outputs probabilities the company calls 'calibrated decisions,' making it fast, cheap (output tokens free, input metered by the billion), and hallucination-free since users define outputs in advance. Almeida says the 'System One model' is trained exclusively on synthetic data via a technique he calls 'reinforcement learning from calibrated decisions,' and demand was high enough to briefly knock out the company's API. Vercel reported 5-18x faster results than ChatGPT Luna 5.6 for command-safety classification, while Bryo AI found Gemini slightly more accurate but 10-20x more expensive for email classification.

TechCrunch · AIupdated · 7h agofirst · 1d agoModel release 2 sources

Appears with

product Jev 3vendor TypeSafe AI 3organization Vercel 2ai model Jev 2vendor Vercel 2organization Bryo AI 1organization Convai Innovations 1organization Earendil 1organization MarkTechPost 1organization OpenAI 1product ChatGPT Luna 5.6 1ai model ChatGPT Luna 5.6 1product GPT-5.6 Terra 1product Laya 1product mmBERT 1product ModernBERT 1
Entities are extracted by the model from each article. Watching an entity keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.