MemoryPlugin
One shared memory across ChatGPT, Claude, and 21+ AI tools
- Category
- Chatbots & Assistants
- Pricing
- 7-day free trial; Core $89/yr, Pro $180/yr, Lifetime $400 one-time
- Best for
- MemoryPlugin fits people who regularly bounce between multiple AI assistants (say, ChatGPT for one task, Claude for coding, Gemini for research) and are tired of re-explaining context, as well as coding-focused users who want Claude Code, Cursor, and Codex sessions folded into the same memory.
- Official site
- www.memoryplugin.com
- Last updated
- August 2026
MemoryPlugin is a third-party memory layer for AI chat tools, built to solve a specific problem: ChatGPT, Claude, Gemini, and other assistants each maintain their own separate, siloed memory, so context typed into one tool never reaches another. It works by capturing memories a user (or the AI) flags as worth keeping, importing existing chat history from platforms like ChatGPT and Claude, and letting users upload reference files, then injecting that context automatically into future conversations across whichever AI tool they're using. Access happens through a browser extension (Chrome and Safari, including mobile via Quetta Browser), a hosted MCP server for developer tools like Claude Code, Cursor, and Windsurf, a Custom GPT for ChatGPT's mobile and desktop apps, a TypingMind plugin, and an open API, giving it one of the widest integration footprints of any AI memory product.
The product organizes memory into two layers: a curated "notebook" of discrete, user-controlled memories grouped into buckets (work, personal, projects), and an "archive" of imported chat history that can be searched by meaning across years of past conversations. A feature called Smart Memory claims to inject only the relevant subset of a user's memory into each conversation, reducing token usage by up to 90% compared to dumping raw history into every prompt. MemoryPlugin positions itself against both native, built-in AI memory (which it argues fills up, gets silently rewritten into vague summaries, and stays locked inside one app) and developer-oriented alternatives like Supermemory, OpenMemory, and Basic Memory, which it describes as requiring self-hosting or manual Markdown management rather than being a ready-to-use consumer product. A free macOS app, MemoryPlugin Sync, additionally pulls Claude Code, Codex, and Cursor sessions into the same searchable archive.
MemoryPlugin fits people who regularly bounce between multiple AI assistants (say, ChatGPT for one task, Claude for coding, Gemini for research) and are tired of re-explaining context, as well as coding-focused users who want Claude Code, Cursor, and Codex sessions folded into the same memory. It's a weaker fit for someone who only ever uses a single AI tool with adequate native memory already (e.g., exclusively ChatGPT Plus), or for privacy-conscious users uncomfortable with a third party storing an importable archive of their chat history.
Key features
Cross-Platform Memory Sync
One memory store shared across ChatGPT, Claude, Gemini, Grok, Perplexity, DeepSeek, Mistral, and 21+ other AI tools, so information entered in one tool is available in the next.
Buckets & Curated Memory
Memories are organized into buckets (e.g., work, personal, a specific project) that can be shared with teammates on the Pro plan, keeping context separated by purpose.
Chat History Import
Imports existing conversation history from ChatGPT, Claude, TypingMind, and Grok via direct file export, and from Gemini and DeepSeek via ongoing browser-extension sync, making years of past chats searchable.
Smart Memory
Automatically selects and injects only the relevant memories for a given conversation instead of the full history, which MemoryPlugin claims cuts injected tokens by up to 90%.
MCP Server & Agent Skills
A hosted (or local) Model Context Protocol server connects memory to Claude Code, Cursor, Windsurf, Claude Desktop, and 100+ other MCP-compatible clients, plus an open-source agent skill that teaches coding agents to recall and save context automatically.
Browser Extension & Custom GPT
A Chrome/Safari extension injects memory directly into web-based AI chat interfaces, while a Custom GPT provides the fastest path into ChatGPT's mobile and desktop apps.
File Q&A
Users can upload PDF and Word files and ask questions across memories, files, and chat history in one place, with cited sources in the answer.
Version History & Memory Suggestions
The system flags likely duplicate or stale memories (e.g., an outdated city or preference) for the user to merge or update, keeping the memory store accurate over time.
Pricing breakdown
Core
- Unlimited memories across 21+ AI tools
- Buckets, memory suggestions, and Smart Memory
- File Q&A, image memories, version history
- Chat history import: last 500 conversations
Pro
- Everything in Core
- Unlimited chat history import
- Knowledge graph of everything discussed (early beta)
- Conversation summaries and monthly insights
- Usage analytics and shared team buckets
Lifetime
- Permanent access to Pro-tier features
- No recurring billing
Pros and cons
Pros
- Genuinely cross-platform — the browser extension, MCP server, Custom GPT, and TypingMind plugin cover 21+ tools including ChatGPT, Claude, Gemini, Grok, Perplexity, and coding agents like Claude Code and Cursor, so memory isn't locked to one vendor's ecosystem.
- Chat history import (from ChatGPT, Claude, Gemini, Grok, DeepSeek, and TypingMind) turns years of old conversations into searchable, retrievable context rather than losing them to app scrollback.
- Smart Memory claims up to 90% fewer tokens injected per conversation by curating what's relevant, keeping context costs down compared to injecting raw history into every prompt.
- Memories are discrete, user-editable, and deletable rather than being auto-rewritten into an opaque rolling summary, which is how ChatGPT's and Claude's native memory currently work.
- Multiple integration surfaces (browser extension, hosted MCP server, Custom GPT, TypingMind plugin, open API) mean it fits a wide range of workflows without forcing one "correct" client.
- A 7-day free trial paired with a 14-day refund policy lowers the risk of testing the product before committing to annual billing.
Cons
- Pricing is presented almost entirely as annual subscriptions ($89-$180/year) or a $400 lifetime buy-in; there's no visible month-to-month option for someone who only wants to try it briefly at low commitment.
- The Core plan's chat history import caps at the last 500 conversations, so heavy users need to upgrade to Pro to get full historical import.
- The product has no independent presence on G2, Capterra, or Trustpilot as of this research — the "4.5 stars from 100+ reviews" figure shown on MemoryPlugin's own homepage is a self-reported aggregate rather than third-party verified data.
- Because it stores full chat history (once imported) and memory summaries in the cloud, privacy-conscious users are trusting a third party with a searchable archive of years of AI conversations, even with encryption-at-rest claims.
- The knowledge graph feature and team-sharing tools on the Pro tier are explicitly labeled early beta, so more advanced use cases are still maturing.
Alternatives to MemoryPlugin
ChatGPT
OpenAI's general-purpose AI assistant
Compare →Claude
Anthropic's AI assistant, known for careful, high-quality writing
Compare →Perplexity
AI answer engine with real-time, cited web search
Compare →Google Gemini
Google's AI assistant integrated across Workspace
Compare →Dapta AI
All-in-one AI voice and text agents built for SMB sales automation
Compare →Leadster
The conversational marketing chatbot built to turn website traffic into qualified leads.
Compare →NewMode (Algomo)
Turns anonymous website traffic into identified, qualified sales pipeline.
Compare →Poe
One subscription, every major AI chatbot in one app
Compare →HARPA AI
Turn your browser into an AI automation agent
Compare →Bardeen
AI automation that scrapes the web and fills your pipeline
Compare →superwhisper
Speak naturally, get polished text in any app
Compare →Frequently asked questions
Does MemoryPlugin replace ChatGPT's or Claude's built-in memory?
No, it operates alongside native memory as a separate layer accessed via browser extension, MCP server, or Custom GPT, and it explicitly supports importing/exporting around changes to ChatGPT's own memory system.
Is there a free plan?
There's no permanent free tier; every subscription starts with a 7-day free trial that requires a card on file, and cancelling before day 7 avoids any charge.
Can it import my old ChatGPT or Claude conversations?
Yes, direct file export import is supported for ChatGPT, Claude, TypingMind, and Grok, with ongoing online sync for Gemini and DeepSeek via the browser extension.
Does it work with coding tools like Cursor or Claude Code?
Yes, via the hosted MCP server for live use during coding sessions, plus a free macOS app (MemoryPlugin Sync) that imports Claude Code, Codex, and Cursor session history into the archive.
Is my data used to train AI models?
According to MemoryPlugin's privacy claims, no — data is described as encrypted in transit and at rest, kept private, never sold, and never used to train models.
What's the difference between Core and Pro?
Core covers unlimited memories and the last 500 imported conversations; Pro adds unlimited chat history import, an early-beta knowledge graph, monthly insights/summaries, usage analytics, and shared team buckets.
Ready to try MemoryPlugin?
Head to the official site to explore pricing and start a free trial where available.
Visit MemoryPlugin →