Custom GPT

Use MemoryPlugin with OpenAI’s Custom GPT system

Written by asad

Last updated 2 months ago

Custom GPTs are plugins for the ChatGPT app that can be used on mobile, desktop, and web. MemoryPlugin is offered as a Custom GPT that automatically manages your memories.

MemoryPlugin Custom GPT

MemoryPlugin is available as an official Custom GPT:

MemoryPlugin Custom GPT

Access the official MemoryPlugin Custom GPT

Account email independence: Your MemoryPlugin sign‑in email does not need to match the email you use on ChatGPT, Claude, TypingMind, or any other tool. Access and permissions are tied to your MemoryPlugin account (and API key when applicable), not third‑party logins.

How to Use

When using MemoryPlugin via the Custom GPT, you need to start chats with the Custom GPT for it to automatically add and load memories.

Custom GPTs only support GPT-4o and GPT-5 Auto models. If you want to use other ChatGPT models like GPT-5 Pro, GPT-4.1, or legacy models, the browser extension is a better choice.

Starting a New Chat

1. Navigate to the MemoryPlugin Custom GPT
2. Start a new conversation - memories will load automatically
3. The GPT will fetch your existing memories at the beginning of each chat
4. You can ask it to load memories from a specific bucket if needed

Using in Existing Chats

On web and desktop, you can invoke the Custom GPT into an existing chat with certain models:

1. Type @ in the chat input
2. Type “MemoryPlugin” to find and select it
3. Once selected, all messages will be processed by the Custom GPT’s model (GPT-4o or GPT-5 Auto) instead of your chosen model
4. You can ask it to load memories or add new memories while active

Features

Automatic Memory Loading

The Custom GPT is instructed to automatically load memories at the start of new chats, providing immediate context about you and your previous conversations.

Memory Management

• Add memories: Ask the GPT to remember specific information
• Load from buckets: Request memories from specific buckets

Cross-Platform Sync

Memories created through the Custom GPT are synchronized with all other MemoryPlugin integrations, so they’re available in the browser extension, MCP server, and other integrations.

Limitations

Keep these limitations in mind when using the Custom GPT:

• Only works with GPT-4o and GPT-5 Auto models
• Cannot switch to other ChatGPT models while the Custom GPT is active
• Requires starting new chats with the Custom GPT for automatic memory loading
• When invoked with @ in existing chats, all subsequent messages use the Custom GPT’s model

Use Cases

The Custom GPT integration is ideal for:

• Mobile users who want memory functionality on their phones
• Quick memory access without needing to install extensions
• GPT-4o workflows where the model restriction isn’t an issue
• Automatic memory management with minimal setup required

Getting Started

1. Visit Custom GPT

   Go to the MemoryPlugin Custom GPT

2. Start Conversation

   Begin a new chat - the GPT will automatically load your memories

3. Set Bucket (Optional)

   Ask to load memories from a specific bucket: “Load memories from my Work bucket”

4. Use Normally

   Chat normally - the GPT will automatically save important information

Creating Your Own Custom GPT

Want to integrate MemoryPlugin into your own Custom GPT? See our guide on integrating into your own Custom GPT for detailed instructions.

Next Steps

Memory Buckets

Learn how to organize memories into different contexts

Browser Extension

Try the browser extension for more model flexibility