# Language Model Info

In any space or module, the **language model** is configured as described in this page.

For **Azure** and **LiteLLM-deployed models**, only the **model name** is required. For **custom-deployed models**, additional parameters must be provided in **JSON format** to fully define the model configuration.

# Overview

The **Language Model Info** schema is used to configure different language models within our tool.

For **Azure OpenAI models** or models deployed via **LiteLLM**, setup requires only the model name, no additional configuration is needed. For **custom models**, a full configuration is required as a JSON, as outlined in the sections below.

# Example Configuration

## Azure Model Configuration

For Azure models, a simple configuration can be achieved by specifying the model name.

```json
{
  "language_model": "AZURE_GPT_4_0613"
}
```

## Custom Model Configuration

Custom models require a full configuration setup, detailing all necessary parameters:

```json
{
  "language_model": {
    "name": "Custom_Model_Name",
    "version": "1.0",
    "provider": "CUSTOM",
    "encoder_name": "o200k_base",
    "token_limits": {
      "token_limit_input": 5000,
      "token_limit_output": 1500
    },
    "capabilities": ["function_calling", "streaming"],
    "info_cutoff_at": "2023-10-01",
    "published_at": "2023-09-01",
    "retirement_at": null,
    "deprecated_at": null,
    "retirement_text": null
  }
}
```

## Fields Documentation

The following table describes the different parameters and which ones are required.

| **Field Name** | **Description** | **Type** | **Default Value** | **Required** |
| --- | --- | --- | --- | --- |
| `name` | The name of the language model, can be an Azure model name or a custom name | string | N/A | Yes |
| `version` | The version of the language model | string | N/A | Yes |
| `provider` | The provider of the language model, either AZURE or CUSTOM | string | N/A | Yes |
| `encoder_name` | The encoder name used for the model | string [enum] | `cl100k_base` | No |
| `token_limits` | Defines input and output token limits for the model | integer | `{7000, 1000}` | No |
| `capabilities` | Lists the capabilities of the model. See below for available capabilities. | array | `[{\"streaming\"}]` | No |
| `info_cutoff_at` | Date when information cutoff occurs | string | `null` | No |
| `published_at` | Date when the model was published | string | `null` | No |
| `retirement_at` | Date when the model is retired | string | `null` | No |
| `deprecated_at` | Date when the model is deprecated | string | `null` | No |
| `retirement_text` | Text describing the retirement details | string | `null` | No |

## Nested Types

### token_limits

| **Field Name** | **Type** | **Description** | **Default Value** | **Required** |
| --- | --- | --- | --- | --- |
| token_limit_input | integer | Maximum number of input tokens | 7000 | Yes |
| token_limit_output | integer | Maximum number of output tokens | 1000 | Yes |

## Enums and Constants

Enumerated types and constants define fixed values to standardize data inputs.

- **Enum Name**: The identifier for the enumeration.
- **Possible Values**: All potential values the enum can represent.
- **Description**: Explanation of the enum's role.

| **Enum Name** | **Possible Values** | **Description** |
| --- | --- | --- |
| EncoderName | `o200k_base`, `cl100k_base`, `qwen`, `deepseek` | Defines the encoder used for the model |
| LanguageModelName | see [LLM Availability Overview](https://docs.unique.ai/administrators/models-llm-management/how-we-evaluate-models/llm-availability-overview) | Lists available Azure language models |
| LanguageModelProvider | `AZURE`, `CUSTOM` | Specifies the provider of the language model |
| ModelCapabilities | `"function_calling", "parallel_function_calling", "reproducible_output", "structured_output", "vision", "streaming", "reasoning"` | Enumerates the capabilities that the model can support |

## Full Json Schema

```json
{
  "$defs": {
    "EncoderName": {
      "enum": ["o200k_base","cl100k_base"],
      "title": "EncoderName",
      "type": "string"
    },
    "LanguageModelName": {
      "enum": ["AZURE_GPT_35_TURBO_0125", "AZURE_GPT_4o_2024_0513", "AZURE_GPT_4o_2024_0806", "AZURE_GPT_4o_MINI_2024_0718", "AZURE_o1_PREVIEW_2024_0912", "AZURE_o1_2024_1217", "AZURE_o1_MINI_2024_0912", "AZURE_o3_MINI_2025_0131", "AZURE_GPT_45_PREVIEW_2025_0227"],
      "title": "LanguageModelName",
      "type": "string"
    },
    "LanguageModelProvider": {
      "enum": ["AZURE", "CUSTOM"],
      "title": "LanguageModelProvider",
      "type": "string"
    },
    "LanguageModelTokenLimits": {
      "properties": {
        "token_limit_input": {
          "title": "Token Limit Input",
          "type": "integer"
        },
        "token_limit_output": {
          "title": "Token Limit Output",
          "type": "integer"
        }
      },
      "required": ["token_limit_input", "token_limit_output"],
      "title": "LanguageModelTokenLimits",
      "type": "object"
    },
    "ModelCapabilities": {
      "enum": ["function_calling", "parallel_function_calling", "reproducible_output", "structured_output", "vision", "streaming", "reasoning"],
      "title": "ModelCapabilities",
      "type": "string"
    }
  },
  "properties": {
    "name": {
      "anyOf": [{"$ref": "#/$defs/LanguageModelName"}, {"type": "string"}],
      "title": "Name"
    },
    "version": {
      "title": "Version",
      "type": "string"
    },
    "provider": {"$ref": "#/$defs/LanguageModelProvider"},
    "encoder_name": {"$ref": "#/$defs/EncoderName", "default": "cl100k_base"},
    "token_limits": {"$ref": "#/$defs/LanguageModelTokenLimits", "default": {"token_limit_input": 7000, "token_limit_output": 1000}},
    "capabilities": {"default": ["streaming"], "items": {"$ref": "#/$defs/ModelCapabilities"}, "title": "Capabilities", "type": "array"},
    "info_cutoff_at": {"anyOf": [{"format": "date", "type": "string"}, {"type": "null"}], "default": null, "title": "Info Cutoff At"},
    "published_at": {"anyOf": [{"format": "date", "type": "string"}, {"type": "null"}], "default": null, "title": "Published At"},
    "retirement_at": {"anyOf": [{"format": "date", "type": "string"}, {"type": "null"}], "default": null, "title": "Retirement At"},
    "deprecated_at": {"anyOf": [{"format": "date", "type": "string"}, {"type": "null"}], "default": null, "title": "Deprecated At"},
    "retirement_text": {"anyOf": [{"type": "string"}, {"type": "null"}], "default": null, "title": "Retirement Text"}
  },
  "required": ["name", "version", "provider"],
  "title": "LanguageModelInfo",
  "type": "object"
}
```