Question 1

What is heyjev.ai?

Accepted Answer

A small, independent playground for Jev, TypeSafe AI’s System One model. Its first experiment, “Hey Jev, should I deploy?”, turns a plain-English description of a deploy into a humorous verdict backed by real probabilities. It is not an official TypeSafe product.

Question 2

How does “Hey Jev, should I deploy?” work?

Accepted Answer

Your text is sent to TypeSafe’s API together with 18 typed questions (yes/no, multiple choice and scored levels). Jev returns a probability for every possible answer in a single request, usually in a few hundred milliseconds, and generates no text. Code then adds up what each answer means for a risk score and maps it to a verdict such as SHIP IT, MAYBE MONDAY or GO TOUCH GRASS.

Question 3

Is it a real deployment gate?

Accepted Answer

No. It is entertainment and education. Do not use it to decide, approve, block or automate a real release, and do not build anything that depends on it: there is no SLA and it can change or disappear at any time. See the terms.

Question 4

Can I call it from my own tools or an AI agent?

Accepted Answer

Yes, for personal non-commercial use within the published limits: POST /api/ask with a scenario of up to 400 characters returns the verdict, the reasons, the raw probabilities, the latency and a share link. A skill file for coding agents is at /skill, and every response carries a notice that it is not advice.

Question 5

What is Jev?

Accepted Answer

Jev is TypeSafe AI’s System One model. Instead of writing text, it takes a state (text or JSON) and typed questions and returns calibrated probabilities for all of them at once, so software can branch on the numbers.

Hey Jev.

Hey Jev, should I deploy?

Ask from your LLM

Playground