Jev
Jev is a small, fast AI model from TypeSafe AI that answers a fixed set of question types with typed values instead of generated text. TypeSafe calls it the first “System One” model, a term borrowed from the idea of quick, intuitive judgment, and positions it for decisions that software consumes directly rather than for open-ended prompts.
Jev exposes three question primitives. A Noul asks a yes/no question and returns the probability that the answer is yes, where a value near 0.5 means the model is undecided. A Choice picks one option from a set of up to 255 and returns the chosen key, a probability per option, and a confidence value. A Score places an input on an ordered rubric and returns a probability-weighted position that can land between two levels.
Because every answer is a number or a key, a program can branch on it without parsing prose, which makes Jev an alternative to coaxing structured output from a general-purpose LLM.
TypeSafe launched Jev in September 2026 with sub-second latency and per-token input pricing far below frontier models. Access goes through the TypeSafe API or through OpenRouter, which serves the jev-latest alias and pinned versions such as jev-1.13.
In Python, the official typesafe-sdk package on PyPI provides sync and async clients. Questions are built from Noul, Choice, and Score objects, sent together in a single request, and returned as validated Pydantic models. Only the early SDK releases exist so far, and minor versions have already introduced breaking changes to the question types.
Official website: typesafe.ai
Related Resources
Tutorial
How to Use the OpenRouter API to Access Multiple AI Models via Python
Access models from popular AI providers in Python through OpenRouter's unified API with smart routing, fallbacks, and cost controls.
For additional information on related topics, take a look at the following resources:
- Accessing Multiple AI Models With the OpenRouter API (Course)
- Pydantic AI: Build Type-Safe LLM Agents in Python (Tutorial)
- How to Use the OpenRouter API to Access Multiple AI Models via Python (Quiz)
- Building Type-Safe LLM Agents With Pydantic AI (Course)
- Pydantic AI: Build Type-Safe LLM Agents in Python (Quiz)
By Philipp Acsany • Updated Sept. 23, 2026
