> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vainona.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# jev current

> Limits, data handling and conformance results for jev current.

<Info>Conformance run pending. This version becomes `active`, and `GET /engines` lists it, once it passes the conformance suite.</Info>

Use this version with `"engine": {"name": "jev", "version": "current"}`. `current` runs whichever Jev version the provider serves. Every answer records the epoch of model behaviour that produced it; a new epoch begins only when we detect a change, and then we re-check quality against the conformance suite.

## Limits

|                       |                                                       |
| --------------------- | ----------------------------------------------------- |
| Context               | 32,000 tokens for the state plus the longest question |
| Whole request         | 64,000 tokens                                         |
| Choice options        | 255, including the `none_of_the_above` escape option  |
| Score levels          | 2 to 10                                               |
| Questions per request | 32                                                    |
| Tokenizer             | approximate (`cl100k_base-approximation`)             |

Judging on this version is billed in [judgment units](/pricing#judgments), the same as on every engine.

## Data handling

|                    |                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     |
| ------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Hosted by          | TypeSafe AI, in the US                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                              |
| Region             | us                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  |
| Subprocessor       | TypeSafe AI operates the model. The full list of subprocessors is in the data processing agreement.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 |
| Customer data      | The compiled context is sent to the engine provider's API with zero data retention requested on every call.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         |
| Version retirement | Version `current` is not pinned: it is whatever model the provider serves now, and that can change without notice. Every answer records the epoch of model behaviour that produced it: `engine_version` is `current+` the date the epoch began and its number, such as `current+2026-09-24.1`, or `current+pending` in the first minutes after a new deployment. A new epoch begins only when we detect a change. The model's answers vary slightly between identical calls, so we probe it every 30 seconds and compare 10-minute averages; when its behaviour changes we start a new epoch, with the next number, and every answer from then on records it. Answers already computed keep the epoch they were computed under. `current` itself is never retired: a change of model starts a new epoch instead. If the provider stopped serving Jev altogether, judgments on it would fail with `engine_version_unavailable`; we never re-pin a judgment silently. |

## Conformance

No conformance run is recorded for this version yet. Every version passes the same suite before it becomes `active`: 500 human-labelled items per type from a fixed golden set that no engine is tuned to. This page shows the results, with the date and the golden set version, once the run exists.