# Model

> **Class** in `langchain_nvidia_ai_endpoints`

📖 [View in docs](https://reference.langchain.com/python/langchain-nvidia-ai-endpoints/_statics/Model)

Model information.

## Signature

```python
Model()
```

## Description

**Thinking mode can be enabled via two mechanisms:**

1. Tag-based: Use thinking_prefix/no_thinking_prefix (appended to
    system message)
2. Param-based: Use thinking_param_enable/thinking_param_disable
    (merged into request params)

All aliases are deprecated and will trigger a warning when used.

## Extends

- `BaseModel`

## Properties

- `model_config`
- `id`
- `model_type`
- `client`
- `endpoint`
- `aliases`
- `supports_tools`
- `supports_structured_output`
- `supports_thinking`
- `thinking_prefix`
- `no_thinking_prefix`
- `thinking_param_enable`
- `thinking_param_disable`
- `base_model`

## Methods

- [`validate_client()`](https://reference.langchain.com/python/langchain-nvidia-ai-endpoints/_statics/Model/validate_client)
- [`validate_thinking_config()`](https://reference.langchain.com/python/langchain-nvidia-ai-endpoints/_statics/Model/validate_thinking_config)

---

[View source on GitHub](https://github.com/langchain-ai/langchain-nvidia/blob/0ca4191cf2239c4990e4706ac1005463f27ff325/libs/ai-endpoints/langchain_nvidia_ai_endpoints/_statics.py#L8)