Jev 1.14 is the second public version of Jev, the System One decision model from TypeSafe AI. Like Jev 1.13, it does not generate text: it returns typed answers — a yes/no with a probability, a choice from a defined set, or a score — with calibrated probabilities for routing, classification and other decision points in an application. It was released on 14 October 2026 and is available on AI/ML API as typesafe/jev-1.14.
box, asks where something is in an image.typesafe/jev Aliastypesafe/jev-1.14 always runs version 1.14. The alias typesafe/jev points to the newest Jev release and resolves to 1.14 today, but will move when TypeSafe ships the next version. Use the pinned id when probabilities feed fixed thresholds. Jev 1.14 is also the decision model behind the Jev Router on AI/ML API.
noul (yes/no with probability), choice, score, boxPOST /v1/decisionsOnly input is billed, at $0.0546 per 1M tokens — the same rate as Jev 1.13. Jev does not generate text, so there are no output tokens to charge.
Routing requests between models or handlers, as in AI/ML API's Jev Router, which uses Jev 1.14 to pick the bucket a chat request belongs to.
Classification and moderation in multilingual products, where TypeSafe reports the largest improvement over 1.13.
Decisions about images: yes/no or choice questions about what an image shows, and Box questions that locate an object in the image.
Scoring content or requests on a defined scale and acting on the score in code, without parsing generated text.
import requests r = requests.post( "https://api.aimlapi.com/v1/decisions", headers={"Authorization": "Bearer " + AIMLAPI_KEY}, json={ "model": "typesafe/jev-1.14", "state": "Help! My payments have been failing for 3 days.", "questions": { "urgent": { "type": "noul", "instructions": "Is this an urgent support request?" } } }, ) print(r.json())
const r = await fetch("https://api.aimlapi.com/v1/decisions", { method: "POST", headers: { Authorization: `Bearer ${process.env.AIMLAPI_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ "model": "typesafe/jev-1.14", "state": "Help! My payments have been failing for 3 days.", "questions": { "urgent": { "type": "noul", "instructions": "Is this an urgent support request?" } } }), }); console.log(await r.json());
curl -X POST https://api.aimlapi.com/v1/decisions \ -H "Authorization: Bearer $AIMLAPI_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"typesafe/jev-1.14","state":"Help! My payments have been failing for 3 days.","questions":{"urgent":{"type":"noul","instructions":"Is this an urgent support request?"}}}'
OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Input | |
| Output | |
Only input is billed, at $0.0546 / 1M tokens. Jev returns decisions rather than generated text, so output is free.
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
Jev 1.14 This page | Typed decisions, multilingual and images | |||
| Reproducible typed decisions in code | ||||
| Typed decisions inside software | ||||
| Cutting LLM spend without picking a model |
Yes. Jev 1.14 is live on AI/ML API as typesafe/jev-1.14, served through the POST /v1/decisions endpoint.
TypeSafe describes 1.14 as better across the board, with the biggest gains in languages other than English. It also extends experimental image input and adds Box questions, which ask where something is in an image.
typesafe/jev-1.14 is pinned to version 1.14. typesafe/jev is an alias for the newest Jev release; it points to 1.14 today and will move to the next version when TypeSafe ships one.
Yes/no with a probability (noul), a choice from a defined set (choice), a score (score) and, new in 1.14, Box questions (box) that locate something in an image.
Input is billed at $0.0546 / 1M tokens, the same as Jev 1.13. Output is free, because Jev returns decisions rather than generated text.
262,144 tokens.
