Anthropic adds 'Memory' feature to Claude AI, helping to remember and maintain conversation context like ChatGPT
After a long wait, Anthropic has officially brought the 'Memory' feature to the Claude AI model, helping this model catch up with big competitors like ChatGPT and Google Gemini in the personalized AI race.
The feature was initially only available to Claude Team and Enterprise users, but starting this month, Anthropic is expanding it to paid personal users, starting with the Max and Pro plans, rolling out gradually over the next two weeks.
Claude can now 'remember' you
Previously, every time a user started a new conversation, Claude would 'forget' everything that had been exchanged before. But with Claude Memory, this AI assistant can now remember the work content, usage habits and project context, thereby maintaining the conversation thread and significantly increasing work efficiency.
The most interesting thing is that Claude does not store the entire memory, but creates separate 'memory spaces' for each project. That is, each project has its own independent storage area, which helps to separate the content and avoid confusion between different tasks.
Unlike ChatGPT, which tends to retrieve information from the entire chat history, Claude only retrieves data from relevant 'memory regions'. Users can also tell Claude to prioritize certain memories or even tell it to 'forget' some past data when necessary.
Additionally, Anthropic adds an Incognito chat mode — where Claude doesn't save any content to memory or history, and all data disappears as soon as you close the chat session.
Added new skills for Claude
Along with the memory feature, Anthropic also launched Claude Skills — allowing users to create custom instructions as Markdown files, for Claude to perform specific tasks like analyzing data, creating reports, or automatically generating documents.
The company also just introduced Claude Code on the web and iOS, allowing users to push code, create pull requests, and many other programming operations directly from the browser or phone.
How useful is memory really?
With 'memory', Claude can understand context across multiple sessions. Users can ask questions like:
'What was I doing last week?'
Claude will instantly reiterate work progress, summarize the project, or suggest next steps.
Users can also import memo data from other AI tools like ChatGPT, or export memo from Claude to another platform, depending on their information management and security needs.
With the addition of Claude Memory, Anthropic closes the gap with ChatGPT and Gemini, bringing its AI assistant closer to being able to work continuously, intelligently, and more personalized.
You should read it
- Anthropic Launches Claude 2: New Competitor for ChatGPT and Bard
- Anthropic Announces Claude Opus 4, the World's Most Powerful Programming Model
- How to use Anthropic's new AI Claude 3 Prompt Library
- Anthropic launches Claude 3.5 Sonnet, beating ChatGPT 4o
- Compare Claude 3.5 Sonnet, ChatGPT 4o and Gemini 1.5 Pro
- Anthropic Launches Claude Haiku 4.5 – Its Fastest, Most Economical, and Safest AI Model Ever