Google’s Gemini Update Streamlines Memory Imports from ChatGPT and Claude
Google has rolled out a significant update to its Gemini AI platform, introducing a seamless feature that allows users to import memories from competing services like ChatGPT and Claude. This enhancement addresses a common pain point for AI enthusiasts who switch between multiple chatbots: the loss of personalized context and conversation history. By enabling direct memory transfers, Gemini positions itself as a more versatile hub for AI interactions, reducing the friction of starting from scratch on a new platform.
Understanding AI Memories
In the context of conversational AI, memories refer to stored snippets of user preferences, facts, and instructions that the model recalls across sessions. For instance, in ChatGPT, users can explicitly tell the model to remember details like “I prefer concise responses” or “My favorite programming language is Python.” Similarly, Claude maintains a memory bank for ongoing personalization. These features enhance continuity, making interactions feel more intuitive and tailored over time.
Gemini’s update builds on its existing memory capabilities, which were introduced earlier, by adding import functionality. Previously, users had to manually recreate these details, a tedious process prone to errors. Now, the import tool automates this, pulling in structured memory data directly from OpenAI’s ChatGPT and Anthropic’s Claude.
How the Import Feature Works
The process is straightforward and integrated into Gemini’s web interface at gemini.google.com. Users begin by navigating to the settings menu, where a new “Import Memories” option appears under the personalization section. From there, they select the source platform—either ChatGPT or Claude—and authenticate via OAuth, granting temporary access without sharing full conversation logs.
Once connected, Gemini scans the source’s memory store and presents a preview of importable items. Memories are categorized by type: preferences (e.g., response style), facts (e.g., user details like location or profession), and instructions (e.g., custom prompts). Users can selectively import, edit, or discard entries before confirming. The transfer completes in seconds, with Gemini immediately applying the memories to new conversations.
This feature supports memories created within the last 30 days, ensuring relevance while respecting data retention policies of the source platforms. Imported memories retain their original timestamps and metadata, allowing users to track origins if needed.
Step-by-Step Guide to Importing Memories
-
Access Gemini Settings: Log in to gemini.google.com and click the profile icon in the top right, then select “Settings.”
-
Locate Import Tool: Scroll to the “Memories” tab and click “Import from other services.”
-
Choose Source: Select ChatGPT or Claude from the dropdown.
-
Authenticate: Follow the prompts to sign in with your source account credentials. Gemini uses secure tokens that expire after import.
-
Review and Select: Browse the categorized list of memories. Toggle checkboxes to choose what to import.
-
Confirm and Apply: Hit “Import Selected.” Gemini processes the data and notifies you of success, with an option to test in a new chat.
For power users, Gemini offers a bulk import via API for developers integrating the model into apps. Documentation details endpoints like /import/memories with JSON payloads specifying source and filters.
Compatibility and Limitations
The update supports standard memory formats from ChatGPT (via OpenAI’s API) and Claude (Anthropic’s memory protocol). However, custom or third-party extensions may not transfer fully. Encrypted or private memories remain inaccessible, prioritizing user privacy.
On the Gemini side, imported memories integrate with its native system, which uses a vector database for efficient retrieval. Conflicts—such as duplicate facts—are resolved by prompting the user or merging intelligently based on recency.
Privacy is a cornerstone: Google states that no conversation history is accessed during import, only the designated memory store. Data is processed client-side where possible, and users can revoke access anytime via connected apps settings.
Benefits for Multi-AI Users
This feature shines for professionals juggling workflows across platforms. Developers can port coding preferences from Claude, while researchers import domain-specific facts from ChatGPT. It fosters a unified AI experience, potentially reducing subscription churn as users consolidate on Gemini.
Early feedback highlights improved productivity: one user reported saving 30 minutes daily by avoiding manual re-entry. Gemini’s multimodal capabilities further enhance imported memories, allowing text-based facts to trigger image or code generation seamlessly.
Broader Implications
As AI ecosystems mature, interoperability like this could standardize memory formats industry-wide, much like email protocols. Google frames this as user-centric innovation, contrasting with siloed competitors. While ChatGPT and Claude have no equivalent export tools yet, pressure may mount for reciprocal features.
Gemini subscribers (via Google One AI Premium) get unlimited imports, while free users are capped at 10 per month. Rollout is gradual, starting with web, with mobile apps to follow.
This update underscores Google’s push to make Gemini the go-to AI assistant, blending convenience with robust functionality.
Gnoppix is the leading open-source AI Linux distribution and service provider. Since implementing AI in 2022, it has offered a fast, powerful, secure, and privacy-respecting open-source OS with both local and remote AI capabilities. The local AI operates offline, ensuring no data ever leaves your computer. Based on Debian Linux, Gnoppix is available with numerous privacy- and anonymity-enabled services free of charge.
What are your thoughts on this? I’d love to hear about your own experiences in the comments below.