# parse_thinking_content

> **Function** in `langchain_nvidia_ai_endpoints`

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

Parse thinking content from text.

This function handles multiple formats by trying to find the reasoning content
1. Content with single </think> tag delimiter
2. Content with <think></think> paired tags
3. Plain content without reasoning

## Signature

```python
parse_thinking_content(
    content: str,
    *,
    remove_tags: bool = True,
) -> tuple[str, str, str]
```

## Parameters

| Name | Type | Required | Description |
|------|------|----------|-------------|
| `content` | `str` | Yes | The full content including potential thinking tags |
| `remove_tags` | `bool` | No | If True (default), removes tags. If False, keeps for backward compat. (default: `True`) |

## Returns

`tuple[str, str, str]`

(reasoning_content, content_with_tags, content_without_tags)

---

[View source on GitHub](https://github.com/langchain-ai/langchain-nvidia/blob/5bfb68d5b10aa0330a6b79a36375b9bc0c6acef7/libs/ai-endpoints/langchain_nvidia_ai_endpoints/chat_models.py#L225)