# Semantic Memory 2.0

Semantic Memory 2.0 helps your AI remember important details from your conversations in a more efficient way. Instead of simply searching past messages and using that information for responses, the system creates “memories” that capture key details from your chat. This allows characters to retain more information over time, making interactions feel more natural and consistent while staying within the memory limits.

Basically, Semantic Memory allows characters to remember more than is possible with Context Memory, in a way that you can tailor and manage however you see fit.

### How Does Semantic Memory Work? <a href="#how-does-semantic-memory-work" id="how-does-semantic-memory-work"></a>

As you chat, the AI summarises important details from past messages into compact memories rather than storing full messages. This makes it possible to keep more relevant information available, allowing characters to remember more about your conversation. When generating responses the AI searches its memories to find the most relevant details instead of just pulling information from all previous messages. If there are more memories than the system can use at once, it priorities the most relevant ones for your current conversation.

For new conversations, the character will generate memories based on the last 8 to 10 messages, capturing details in groups of 4 at a time. If you're continuing an old conversation, the character will process earlier messages when you send your next message. It does this in batches of 60 messages at a time, so the first response may take a little longer to generate while it builds the memory.

### Editing And Deleting Semantic Memories <a href="#editing-and-deleteing-semantic-memories" id="editing-and-deleteing-semantic-memories"></a>

You have full control over how the AI makes use of Semantic Memory through the Memory Manager, allowing you to edit, delete, pin or add memories. Even if you delete or edit messages in the conversation, the associated memory will remain in the Memory Manager until you manually update it.

Managing your AI’s memory is optional, but it provides transparency so you can see exactly what the character remembers and adjust it. This ensures your experience is personalised and consistent.

{% hint style="info" %}
You can learn more about the Memory Manager and how to use it [here](/product-guides/premium-features/memory-manager.md).
{% endhint %}


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.secretmate.ai/product-guides/premium-features/semantic-memory-2.0.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.