Anthropic Just Killed the 'Which Claude Do I Open' Problem
Anthropic announced today that Claude Cowork and Claude chat are becoming one product — just "Claude." If you've ever paused before opening the app, trying to remember whether your task needed the "work mode" version or the regular chat, that decision point is going away.
Here's what's actually changing, and why it matters more than it sounds.
The Problem: Too Many Front Doors
Claude Cowork launched as a separate surface built for multi-step work tasks — the kind of thing that needed project context, file handling, and longer-running execution. Regular Claude chat stayed the conversational front door.
The catch: users had to decide before they started typing which one they needed. Realize halfway through a chat that you actually need Cowork's capabilities? Start over in a different tab. That's friction that has nothing to do with the actual work.
What's Changing
Starting now, and rolling out to Pro and Max plans over the next few weeks, there's just one Claude. The capabilities that used to live behind the Cowork and Design labels — planning, executing, and producing real deliverables — are available in any conversation, at any point, without switching contexts.
Alongside the merge, Anthropic is also folding in three specific tools directly into the chat surface:
- Claude Docs — draft and edit documents inline
- Claude Slides — build presentations without leaving the conversation
- Claude Design — the same design capabilities Cowork offered, now available by default
Anthropic's Boris Cherny confirmed the detail that matters most for anyone who's tried AI-generated slides before: the output isn't locked inside the chat. Presentations you generate can be opened, edited, and exported straight to PowerPoint or PDF — no separate tool, no copy-pasting into a template afterward.
Why This Is a Bigger Deal Than It Sounds
Splitting a product into "chat" and "work" modes is a natural first step when you're shipping fast — you build the agentic capability, ship it as its own thing, and figure out the merge later. The interesting signal here isn't that Anthropic added slides and docs. It's that they decided the mode switch itself was the thing to remove.
That's a bet on a specific idea: that the friction in AI tools isn't "the AI isn't capable enough," it's "I had to know in advance what I was asking for." Most people don't start a task thinking "this needs an agent" versus "this needs a chatbot" — they just have something they want done. Collapsing that decision back into a single surface removes a step users shouldn't have had to think about in the first place.
What This Means If You Use OpenClaw
OpenClaw has never made you choose a mode. There's no "chat Claude" versus "work Claude" split to begin with — every OpenClaw agent session is built around actually doing the task, whether that's writing code, running a workflow, or producing a deliverable you can hand off.
What Anthropic is converging toward — one surface, full capability, real output formats — is the default assumption an agent framework like OpenClaw starts from. You don't pick a mode before you start; you say what you need done, and the agent figures out what that requires.
The Bigger Picture
The AI industry spent the last couple of years shipping "modes" — chat mode, agent mode, work mode, research mode — because it was the fastest way to ship new capability without breaking the old surface. Anthropic folding Cowork back into Claude is a sign that phase is ending. The next competitive question isn't "which mode has the best output" — it's whose agent needs the least explaining before it gets to work.
If you want to see what that looks like without any mode-picking at all, that's what OpenClaw tutorials are built for.