A decision model answers a fixed question by choosing from options you supply, with a probability for each. Here's how they work, what they're for and where they fall short.
Jev is closed and runs only on TypeSafe's servers. A growing set of open-weight models can run on your own hardware. Here's how they compare on accuracy, compatibility and cost.
Three independent benchmarks rank decision models, and each measures something different. Here's what each one tests and how to read the scores.
10/28/2026: Trending stories on Hackernoon today!
The repo went up on 24 October; the blog named it on the 27th. What d1-Omni-2B actually ships with — specs, audio scores, and the gaps nobody has filled.
TypeSafe AI, led by former OpenAI engineer Diogo Almeida, launched Jev, an AI model for structured decision-making. The company is valued at $7.5 billion weeks after its release. Jev offers faster and cheaper alternatives to traditional LLMs for classification tasks.
d1-2B went live on 24 October with no launch post. What its repo actually ships, where it beats d1-3B, and the three artifacts it still lacks.
20 Local Decision Models that run locally free, and most of them before better than the original Jev.
OpenAI is launching its decision model to the world. Can it compete with the original Jev?
Decision models are emerging as a new class of AI built for fast, structured choices. Here’s how they work, why they’re fueling AI agents and why confidence scores matter.
Snowflake Decision Fast brings low-latency AI decision models to Cortex AI. Classify, score and route data in real time with SQL or the REST API.
Compare 20 Jev alternatives from Cloudflare, Perplexity, OpenAI and open-source teams on price, license, hardware and published benchmarks.
TypeSafe launched Jev on September 15, 2026. In about four weeks, at least 20 alternatives shipped — from Cloudflare, OpenAI, Perplexity, Liquid AI, AWS, and independent developers.
AI decision models offer low-cost SOC triage, but live scans found 51% of infected endpoints had already been marked mitigated by EDR.
A Pac-Man-inspired benchmark called JevMan tests a new generation of AI decision models from TypeSafe, Cloudflare and others, revealing why consistent judgment may matter more than speed.
Never making type errors does not mean never making errors
Jev is a decision model, not a text generator: it takes a state and a list of multiple-choice questions and returns a probability for each answer, rather than producing free-form output. Here are four practical coding use cases.
Jev and the flood of AI decision models — TypeSafe allegedly copied researcher's architecture but got copied itself
Learn how to deploy decision models with llama.cpp on Amazon SageMaker AI. This technical how-to serves OpenAI-compatible chat completions with the AWS Deep Learning Container for llama.cpp, then extends the runtime with SemIf to return probabilities over declared options on CPU or GPU.
Jev's rapid rise was no accident. The growing complexity of AI agents has created an urgent need for fast, cheap, and flexible decision-making.
During its recent “Birthday Week”, Cloudflare announced Clef, a set of open-weight AI models designed to choose between predefined options rather than generate text. Cloudflare released 9B- and 27B-parameter models, along with a platform for adapting them to specific decision-making tasks.
Before founding TypeSafe, the company behind Jev, Diogo Almeida worked at OpenAI. Now his former employer wants a piece of his company's success.
On Tuesday, Musubi announced a lightweight decision model made for real-time moderation called PolicyLM-1.7B, released with open weights.
How does a question-answering architecture stack up against traditional classification on content safety tasks? Jev’s been making the rounds lately, and for good reason — instead of generating prose, it answers named questions directly with a......
Amazon Web Services' Strand Labs has released the latest Jevalike decision model, Strands Decider 2B.
Jev is TypeSafe's structured decision model. It returns bounded judgments that application code can use when generated prose is unnecessary — classifying intents, rating AI responses, or deciding document relevance.
Early signs suggest Jev is gaining traction among developers, with a surge in demand last week causing TypeSafe AI to briefly lose the ability to serve users from its API.
TypeSafe AI's Jev offers developers an alternative to general-purpose reasoning models for routing and decision-making in agentic workflows.
Last week saw the debut of TypeSafe AI's Jev, its first "System One" model. Rather than chatting with users like conventional LLMs, it's strictly designed for statement evaluation and decision-making, for programming purposes.
A claim-by-claim audit of TypeSafe AI's 200x faster, 400x cheaper decision model, with real pricing math and Hacker News skeptic pushback.