Anthropic has introduced a major upgrade for Claude. You now get the much-anticipated Memory feature — similar to ChatGPT's memory — which is now available for Team and Enterprise customers. But, with so many rival chatbots already offering the important feature that makes the chatbot "remember," I have a feeling free users might be seeing it soon.
Clause Team and Enterprise users will now automatically have their settings retained, which means the chatbot will remember everything from project details and preferences to the user's name. The feature is optional and can be fully controlled via settings. Users who want to view, edit or disable what the chatbot remembers can easily do so, The option to go completely "incognito" (similar to ChatGPT's Temporary chat) is also still an option.
This much anticipated memory feature comes at a cost. It's not even in Max, yet, which is $100, month. The next cheapest tier is Pro, which is $17/month. Yet, I still think free users might see it sooner. Here's four reasons why I think this feature won't stay limited to paid tier users for long.
Free vs paid memory gap is too competitive
Anthropic’s competitors already offer memory for free. Everyone from OpenAI, Google Gemini and other best chatbots have some form of memory-like features across more tiers. Free, Pro, or Max users often complain about having to constantly re-establish context. As memory becomes expected rather than exceptional, free users will push for access.
Max plan shows Anthropic is already experimenting
The recently launched Max plan (April 20225) reinforces the idea that Anthropic is segmenting features by usage and need. In my mind, that means it's only a matter of time until features initally reserved for Enterprise will trickle down, even if partially for lower tiers, especially if the infrascruture already supports it.
Incognito mode for all users is already live
Memory may be locked behind paid plans right now, but Anthropic made Incognito Chat available for everyone. That shows the company believes features that touch privacy and memory are viable across tiers, even for non-paying users.
In other words, with the appropriate controls, if Anthropic is confident in their privacy-first rollout, Memory is a natural next candidate for all users.
Technical & product stacking
Here's the thing. Anthropic’s Memory is project-scoped, and Memory summaries are editable. Meaning, you can determine what things the chatbot remembers. These features suggest a solid groundwork for broader deployment. They give users great control, help set boundaries for projects vs. general use and safe scaling. Once internal performance, cost and privacy tests pass, expanding access becomes much more feasible.
For all tiers to get the Memory option, Anthropic will need to ensure privacy controls are airtight so free-tier uses feel safe even enabling it.
Other factotrs include performance costs (storage, retrieval, context management), which will need to be optimized so Memory for lower subscription levels doesn’t degrade service.
The takeaway
When this Claude's Memory feature is rolled out for free tier (I say "when" because I'm that confident), the strategy might start with limited Memory length or project-only memory for free/Max users before opening the full feature.
I also imagine there will also be clear opt-in settings, summaries and UI cues, which are essential for users to know what memory is active and how to manage, view, or delete it.
Right now, Memory is a paid teams feature, but everything Anthropic is doing suggests that this isn’t a permanent restriction. As user expectations rise and AI assistants become more central to both work and life, memory seems to me like a standard feature, rather than an extra.
If you use Claude Pro or Free keep an eye out — this feature might be closer than it seems. Of course, when it does roll out, I'll be among the first to share the news.
Follow Tom's Guide on Google News to get our up-to-date news, how-tos, and reviews in your feeds. Make sure to click the Follow button.
