Anthropic has quietly shipped one of the most meaningful quality-of-life upgrades to Claude in months. On July 10, 2026, the company rolled out a significant update to how Claude’s memory system works — and if you use Claude regularly for work, research, or creative projects, you are going to notice the difference almost immediately.
What Changed
Previously, Claude’s memory operated more like a running summary — a blob of text that Claude would reference when needed. The new system is fundamentally different. Memory now works as a set of individual, categorised entries that Claude actively reads and updates during your conversations.
Instead of one long paragraph about who you are and what you like, Claude now maintains distinct, organised memory cards — one for your professional context, one for your communication preferences, one for your ongoing projects, and so on.
The practical result is that Claude can retrieve specific pieces of information about you with much greater precision. If you told Claude three months ago that you prefer responses without bullet points, or that you are working on a product launch in Q3, that information now lives in a dedicated memory entry rather than being buried in a general summary where it might get overlooked.
Why It Matters
The previous memory system worked reasonably well for casual use but started showing cracks when users had complex, multi-domain relationships with Claude.
A power user who discussed cooking, work strategy, personal finance, and creative writing across dozens of conversations would end up with a memory summary that was trying to do too much at once. The categorised entry system solves this by separating context into lanes, making retrieval more reliable and updates more targeted.
Anthropic has also made memory management more transparent with this update. Users can now see their individual memory entries in Settings and edit or delete specific ones without wiping their entire memory history — a frustration many users flagged with the earlier system.
Who Gets It
The updated memory system is available across Free, Pro, Max, Team, and Enterprise plans. Memory must be enabled in your settings for it to work — it is not on by default for all users. If you are on a Free plan and have not yet turned memory on, this update is a good reason to do so.
This memory upgrade also ties directly into a separate feature Anthropic launched the day before — the new Reflect dashboard, covered in the next story — which uses your memory data to generate a monthly recap of how you have been using Claude.
The two features are designed to work together as part of a broader push toward making Claude feel less like a tool you pick up and put down and more like a persistent, context-aware collaborator.
Watch this video for a full walkthrough of how Claude’s memory system works:
Quick Links:
