Claude Launches Memory Feature
Anthropic's Claude AI now has a free memory feature, letting users save data from conversations and even import/export it. The update makes it easier to switch from other LLMs like ChatGPT while retaining context, a huge workflow improvement for developers who rely on AI assistants for coding and research, sparking over 2,100 likes on one reaction post.
Anthropic's memory feature distinguishes itself from ChatGPT's by being less automatic; it doesn't build a persistent background profile of the user. Instead, it functions more like an on-demand tool, searching through past conversations for relevant context only when explicitly asked by the user, a design choice Anthropic frames as a privacy-first approach. For developers, memory operates via a client-side tool that you control, creating, reading, and updating files in a local `/memories` directory. This system uses `CLAUDE.md` files for explicit instructions and an "auto memory" function that saves notes on build commands, workflow habits, and code style preferences, keeping the active context window focused on the current task. The import tool works by generating a specific prompt for you to paste into a competing AI like ChatGPT. This prompt instructs the old AI to export all stored preferences, behavioral instructions, and context about tools and frameworks into a single code block, which is then ingested by Claude as a starting point. It transfers the learned profile, not the raw conversation history. Previously a feature for paid Team and Enterprise subscribers since late 2025, memory was made available to all free users in early March 2026. This move is part of a broader strategy by Anthropic to gain market share by offering pro-level tools for free and committing to an ad-free experience, funded by subscriptions and enterprise contracts. Users have granular control over what is remembered. In the settings, you can view, edit, or completely reset Claude's memory summary. The system also offers an "Incognito" mode for any conversations that a user does not want stored or referenced in the future. This focus on developer workflows aligns with Anthropic's stated long-term goal of automating AI research and software development. By building tools that offer persistent, user-controlled context for complex coding projects, the company is positioning itself to be a foundational tool for engineers and enterprise clients.
