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

# Get Judge Version

> Retrieve one version of a judge's rubric.

### Overview

> [!NOTE]
>
> Dispositions are enabled per organization. If your organization does not have access, these endpoints return `404`.

A judge is an evaluation rubric applied to call evidence. Returns one version of a judge's rubric: its prompts and the levels it can select, with each level flagged as a target or not. Any version of the judge can be read, but a disposition can only pin the published `judgeVersionId` listed by [Get Judge Catalog](https://docs.bland.ai/api-v2/get/agents-id-dispositions-judge-catalog.md).

### Headers

**authorization** `string` *required*

Your API key for authentication.

### Path Parameters

**agent\_id** `string` *required*

The agent's unique identifier. A malformed id returns `400` with the message `agentId must be a valid UUID`.

**judge\_id** `string` *required*

The judge's unique identifier. Must be organization-wide or scoped to this agent; otherwise returns `404` with the message `Judge not found`. A malformed id returns `400` with the message `judgeId must be a valid UUID`.

**version\_id** `string` *required*

The version's unique identifier. Must belong to `judge_id`; otherwise returns `404` with the message `Judge version not found`. A malformed id returns `400` with the message `versionId must be a valid UUID`.

### Response

**data** `object`

The judge version.

**judge version object**

**id** `string`

Unique identifier for the version.

**judgeId** `string`

The judge this version belongs to.

**versionNumber** `integer`

Sequential version number within the judge.

**name** `string`

Display name at this version.

**description** `string | null`

Description at this version.

**modality** `string`

`text` when the judge reads the transcript, `audio` when it listens to the call recording.

**systemPromptMd** `string`

The rubric's system prompt in Markdown. May be empty.

**promptMd** `string`

The rubric's task prompt in Markdown. May be empty on an unpublished draft.

**levels** `array`

The outcome levels the judge chooses from, in rubric order. Each has `key` (the level's stable key), `label` (display name), `promptMd` (the criteria for selecting this level), and `target` (`true` when this level counts as a passing outcome).

**errors** `null | array`

`null` on success, or a list of error objects if the request failed.

```json Response theme={null}
{
  "data": {
    "id": "4420749d-ff1a-4edd-ad8a-4b61f2ae963a",
    "judgeId": "cfd538c3-507f-4976-90f4-db92369ee48b",
    "versionNumber": 2,
    "name": "Caller sentiment",
    "description": "How the caller felt by the end of the conversation.",
    "modality": "text",
    "systemPromptMd": "You review customer service transcripts and classify the caller's closing sentiment.",
    "promptMd": "Read the full transcript. Pick the level that best describes how the caller felt in the final third of the call.",
    "levels": [
      {
        "key": "positive",
        "label": "Positive",
        "promptMd": "The caller thanks the agent, confirms the issue is resolved, or sounds pleased.",
        "target": true
      },
      {
        "key": "neutral",
        "label": "Neutral",
        "promptMd": "The caller is matter-of-fact with no clear positive or negative signal.",
        "target": true
      },
      {
        "key": "negative",
        "label": "Negative",
        "promptMd": "The caller expresses frustration, threatens to cancel, or hangs up mid-sentence.",
        "target": false
      }
    ]
  },
  "errors": null
}
```

```json Not Found theme={null}
{
  "data": null,
  "errors": [
    {
      "error": "NOT_FOUND",
      "message": "Judge version not found"
    }
  ]
}
```

***

Docs for agents: [llms.txt](https://docs.bland.ai/llms.txt)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.