Claude Memory Now Free + Computer Use for All Paid Plans

Anthropic just flipped the switch on two features that change how you'll use Claude AI automation every single day: memory is now free for all users, and Claude can now control your computer screen on paid plans via computer use. These aren't incremental tweaks — they represent a fundamental shift from chat assistant to hands-on digital co-worker.

This article breaks down the most important updates from Claude's six-month release cycle (August 2025–March 2026), what each change means for your real workflow, and how Claude now stacks up against ChatGPT and Gemini.

Claude Memory Goes Free: No More Re-Explaining Yourself Every Session

On March 2, 2026, Anthropic unlocked Claude's memory feature (the ability to store and recall facts about you across entirely separate conversations) for all users, including the free tier. Before this, persistent memory required a paid subscription.

Here's how it works in practice: Claude automatically stores key details from your conversations — your job title, writing style preferences, ongoing projects, recurring questions — and resurfaces them in future chats without prompting. You stop starting from zero every time you open a new session.

• What it stores: Your name, role, communication style, project names, recurring needs
• Where it works: claude.ai web, Claude for iOS, Claude for Android
• You stay in control: Go to Settings → Memory to view, edit, or delete any stored fact at any time
• Incognito mode: A no-memory chat option (like a browser's private window — zero storage, zero tracking) is available on Pro, Max, and Team plans

For comparison: ChatGPT's memory feature still requires a Plus subscription at $20/month. Claude is now offering the same capability to free users — a direct competitive move that makes the free tier meaningfully more useful overnight.

Claude Computer Use: AI Automation That Takes the Wheel on Your Desktop

Launched March 23, 2026 as a research preview on Pro and Max plans, computer use lets Claude autonomously operate your desktop — opening files, running programs, clicking buttons, filling out forms, and navigating websites entirely on its own.

Think of it as handing Claude a mouse and keyboard. You describe a task in plain language, and Claude executes every step on your actual screen in real time. This is categorically different from Claude generating code or giving instructions — it's Claude doing the work.

Real-world tasks computer use can handle today:

• Open a spreadsheet, extract specific numbers, and paste them into a report template
• Search the web for competitive pricing, then compile findings into a formatted document
• Run a local Python script and troubleshoot error messages as they appear on screen
• Navigate a web dashboard and complete repetitive data-entry tasks automatically

Current limitations to know: Computer use is available only on Pro and Max plans (not free or standard Team), and it's explicitly labeled as a "research preview" — meaning it handles most tasks reliably but Anthropic recommends starting with low-stakes, easily reversible workflows before trusting it with anything critical.

Claude vs. ChatGPT vs. Gemini: AI Automation Comparison

Neither ChatGPT nor Google Gemini offers equivalent local desktop screen-control in their standard consumer tiers as of April 2026. OpenAI's Operator product exists but operates in a web-only sandbox, not your actual desktop environment. Claude Pro/Max users are accessing a capability that's genuinely rare among mainstream AI assistants right now.

Six Months of Claude Updates: Everything That Shipped Since August 2025

Computer use and free memory are the headliners, but Anthropic quietly shipped more than 25 major feature updates between August 2025 and March 2026. Here's the full timeline of what actually matters:

• Mar 25, 2026: Interactive apps on iOS/Android — live charts and shareable visual assets built inside conversations
• Mar 23, 2026: Computer use research preview on Pro/Max — autonomous screen control
• Mar 12, 2026: Inline custom charts and diagrams generated directly in chat without any separate tool
• Mar 2, 2026: Memory unlocked for ALL free users
• Feb 17, 2026: Claude Sonnet 4.6 — upgraded coding, agent planning, and a 1-million-token context window (in beta) for processing book-length documents in one shot
• Feb 13, 2026: Analytics API for Enterprise — programmatic access to daily AI usage data per organization
• Feb 12, 2026: Self-serve Enterprise plan sign-up — no sales call required anymore
• Jan 16, 2026: Claude Code (AI-assisted coding tool with vibe coding support) added to all Team Standard seats at no extra cost
• Jan 12, 2026: Health and fitness data integration on mobile — activity patterns, workout trends, sleep quality (US users only)
• Jan 12, 2026: HIPAA-ready Enterprise plans for healthcare organizations handling protected patient data
• Nov 24, 2025: Context window compaction — auto-summarizes earlier messages so conversations can run indefinitely without hitting length limits
• Aug 27, 2025: Code Execution Tool — sandboxed Python execution (actually runs your code in a secure container, not just generates it)

The Healthcare and Enterprise AI Expansion Nobody Is Talking About

Two January 2026 updates flew under the radar but signal exactly where Anthropic is placing its biggest strategic bets for 2026.

HIPAA-ready Enterprise plans make Claude eligible for deployment inside hospitals, clinics, insurers, and medical research firms. HIPAA (the Health Insurance Portability and Accountability Act — the US federal law governing how patient records must be stored and processed) compliance was the final barrier keeping Claude out of most healthcare IT environments. With HIPAA-ready status, Anthropic opened a market that's been closed to most AI vendors due to data privacy requirements.

Health data on mobile lets Claude read your fitness tracker and sleep data on iPhone and Android — with your permission. Claude can then answer questions using your actual personal numbers: "My resting heart rate has been elevated for 5 days and I'm averaging 5.8 hours of sleep — what should I watch for?" instead of generic health advice. This is currently limited to US users only.

On the enterprise productivity side, the Cowork plugin marketplace (launched February 24, 2026) lets Team and Enterprise admins deploy and govern third-party integrations with full admin controls. Pair that with a self-serve sign-up flow, the Analytics API (daily usage reporting per organization), and Claude Code included in Team seats, and Anthropic has quietly built a serious enterprise product layer in under 12 months.

What to Do Right Now — By Claude Plan Type

Based on your subscription, here's the immediate action that makes the most sense:

# Quick-start guide by Claude plan

Free users:
  1. Visit claude.ai → Settings → Memory
  2. Review what Claude has already stored and edit as needed
  3. Memory is active by default — start a new chat and test it

Pro / Max users:
  1. Go to Settings → Experimental Features → enable Computer Use
  2. Start simple: "Open my Downloads folder and list the 5 newest files"
  3. Expand gradually to multi-step tasks as you build confidence

Team plan users:
  1. Claude Code is now included in all Standard seats (since Jan 16, 2026)
  2. No extra purchase needed — activate via your Team admin dashboard

Enterprise / Healthcare:
  1. HIPAA-ready plans at claude.ai/enterprise
  2. Self-serve sign-up now available — no sales demo required

If you're evaluating Claude against other AI assistants for your team, our AI automation productivity guide has a side-by-side breakdown by use case. If you're brand new to Claude, the Claude setup guide walks you through everything in under 10 minutes.

Related Content — Get Started | Guides | More News