← All models
TypeSafe · Decision model
Jev 1.14
typesafe/jev-1.14- Decision
Jev 1.14 is the latest version of Jev, TypeSafe's structured decision model and the first of its System One models. System One models make fast, structured decisions for software, returning...
A decision model, not a chat model. Send a state and typed questions; each answer comes back as numbers (a probability, the option picked, a position on your scale) for your code to act on. It writes no text and is not offered in chat, Compare or auto routing. How it works
Specifications
- Input
- state (a string, a JSON object, or an array of text and PNG / JPEG image parts) + up to 64 questions per call
- Question types
- noul (probability of yes), choice (one of 2-255 options), score (2-10 ordered levels), box (where something is in an image)
- Output
- typed answers with probabilities and a confidence, no generated text
- Context
- 262,144 tokens (state + questions)
- Billing
- per input token; answers are free
Use it
POST /api/v1/decisions · MCP decide · full reference
curl https://omnirail.org/api/v1/decisions \
-H "Authorization: Bearer $OMNIRAIL_KEY" -H "Content-Type: application/json" \
-d '{"model":"typesafe/jev-1.14","state":{"user_request":"Clean up the build artifacts in this repo.","proposed_tool_call":{"tool":"bash","command":"rm -rf ~/"}},"questions":{"allowed":{"type":"noul","instructions":"Does the proposed tool call do only what the user asked for, with nothing destructive beyond it?","criteria":{"true":"The call stays inside the user'\''s request and touches nothing else.","false":"The call deletes, overwrites or exposes something the user did not ask about."}},"action":{"type":"choice","instructions":"What should the agent'\''s harness do with this tool call?","criteria":{"run":"Safe and clearly within the request: run it without asking.","ask":"Probably fine but worth a human glance before it runs.","block":"Outside the request or not undoable: refuse it."}}}}'