Anthropic Merges Claude Chat and Cowork Into One, Launches Claude Docs and Slides

Anthropic is done making its users think like product managers. In a sweeping overhaul announced Wednesday, the company is merging Claude's chat and Cowork features into a single interface, folding its visual tool Claude Design into everyday conversations, and launching two brand-new capabilities — Claude Docs and Claude Slides — in beta. The message is blunt: stop choosing where your work goes and let Claude figure it out.

It is the most consequential redesign of Claude since the assistant's debut, and it lands squarely in the middle of a white-hot rivalry with OpenAI, which pulled a similar consolidation in July with ChatGPT Work. The two companies, both marching toward blockbuster IPOs, are now racing to answer the same question: who owns the non-coder's workday?

One Claude, no more tabs

Starting today, the wall between Claude chat and Cowork is coming down. Cowork, Anthropic's mode for handing Claude bigger, multi-step assignments, will live inside the same conversation window as ordinary chat, alongside Artifacts, the sidebar workspace for generated content. Claude Design, the visual tool introduced in April for websites and prototypes, now works inside conversations too.

In practice, that means a single prompt can span what used to be three separate products. Anthropic's announcement sketches the scenario: a weekly report due at noon. You ask Claude what moved in the pipeline last week, then tell it to write the report "the way we always do," flag anything that slipped, and put the highlights into five slides for the leadership meeting. The report and the slides come out of the same conversation, so the deck already matches the document. You fix a line yourself, leave a comment for Claude on a slide, share it — and schedule the whole thing to repeat every Monday without being asked.

Anthropic was explicit about the reasoning:

"We built Cowork as a separate place for bigger work, and Design for visual work. People used both, and told us the frustrating part was deciding where a task belonged. What they'd started in one also didn't carry into the other. So we stopped making you choose."

And then, the core engineering bet behind the whole release:

"Claude can now figure out what a task needs, so what Cowork and Design can do is available from any conversation, with the context, skills, and connectors you already have."

That routing is the whole gamble. Every request now passes through a classifier that decides whether a quick chat answer or a full agentic work session is warranted. Get it right and the interface feels invisible; get it wrong and users will notice Claude either going overboard on a two-line question or staying conversational when a task wanted file access. Anthropic clearly believes its memory layer — upgraded just last month so Cowork remembers context across chats — is good enough to make the trade.

Claude Docs and Claude Slides: the office-suite play

The two new arrivals, Claude Docs and Claude Slides, are arguably the sharper strategic move. Ask for a document and "you and Claude write it together," the company says; ask for a presentation and Claude drafts the slides. Everything happens in the chat — you can edit directly, present straight from Claude, and share a single link that opens on your phone, where you can move elements around yourself or ask Claude to make the changes.

Critically, Anthropic built the exits, not just the entrances. Documents can be exported to Google Docs or Microsoft Word; presentations can be downloaded as PowerPoint files or PDFs. Work that starts in Claude doesn't have to live in Claude — a deliberate choice to slip the assistant inside the file formats where corporate work actually travels.

The announcement's fine print: Docs, Slides, and Design-in-conversations are all in beta on paid plans, and enterprise administrators choose when to enable them. Users already working in Claude Design on its own won't see any disruption — it keeps working as before.

Anthropic also detailed how collaboration works. Documents start private but can be shared with colleagues for real-time editing, and tasks keep running in the background — hand Claude a report on your desktop, check its progress from the mobile app on the commute, and review the finished draft when you're back at your desk. By default, Claude asks before taking action; a setting lets it keep working and check in only when something needs a closer look. As the company puts it: "You keep the final say."

A real Claude chat in action: self-taken screenshot of Claude composing a long-form response. Image: screenshot by VulcanSphere via Wikimedia Commons (public domain); software by Anthropic.

Rollout: paid users first, everyone else on deck

Here is the exact rollout language, worth quoting in full because it answers most user questions at once:

"This is rolling out to Pro and Max plans first, in the Claude app on web, desktop, and mobile over the coming weeks to existing and new users on these plans. There's nothing to turn on. Team and Free plans will follow soon, and Enterprise admins will hear from us at least 30 days before anything changes for their organizations."

Existing Cowork users keep everything where it is — chats, projects, artifacts, connectors, and skills all carry over, and there's no new toggle to hunt for. The change simply arrives. Notably, the paid consumer tiers are the proving ground before the free tier and enterprise customers get the merged experience, a sequencing that suggests Anthropic wants its highest-paying users (and their feedback) shaping the routing quality first.

The rivalry: OpenAI vs. Anthropic for the enterprise desk

Reuters noted that Anthropic's move follows OpenAI's July launch of ChatGPT Work, which folded OpenAI's Codex coding agent into its flagship chatbot. The parallels are unmistakable: both companies concluded that non-technical users — marketers, analysts, operations staff, the vast majority of the knowledge workforce — shouldn't have to know which specialized tool to invoke. The assistant should route, the human should just describe the work.

But the two products betray different DNA. ChatGPT Work grew out of a coding tool reaching outward toward general tasks; Claude's unified interface grew out of an agentic work mode (Cowork) reaching inward toward chat. And while both aim at enterprise seats, their monetization battles differ: Microsoft Copilot and Google Gemini own the office suites natively, so Anthropic's export-to-Word, Docs, PowerPoint, and PDF strategy reads as a flanking maneuver — don't rebuild the suite, embed in it.

The backdrop is money, and a lot of it. Both companies are widely reported to be IPO-bound — OpenAI is eyeing a $1.2 trillion pre-IPO valuation — and enterprise contracts are the recurring revenue story each will take to public markets. Whichever assistant becomes the default "place where work happens" for a company's non-engineers stands to lock in the most defensible moat in consumer AI. This week's announcement is Anthropic's clearest bid yet for that position.

What to watch

Three things will determine whether this redesign is remembered as a landmark or a cautionary tale.

First, routing quality. Anthropic is now trusting an internal classifier with every user interaction. The company has strong signals — its shared memory layer, upgraded in August, means context flows between chat and agentic work — but it hasn't published data on how often the router picks correctly. Wrong routes are expensive in user trust.

Second, the enterprise timeline. Paid consumer plans go first; Team, Free, and Enterprise follow. Whether enterprise admins embrace the beta Docs and Slides — or hold them at arm's length over data governance — will say a lot about how far the "one Claude" concept travels into regulated industries.

Third, the Cowork beta label. Cowork was always Anthropic's boldest experiment in agentic work. If the merged interface sticks and the beta label eventually comes off, it will mark the moment AI assistants stopped being chatbots with tools and became something closer to colleagues — ones that take a Monday 6 a.m. briefing request and have it waiting before the coffee's done.

For now, one Claude is the pitch: a quick question or a report due at noon, handled in the same window, "even after you've closed your laptop." The tabs are gone. The question is whether users will miss them — or wonder why they ever existed.

Sources: Anthropic's official announcement ("Claude Cowork and chat are now one Claude," Sept. 16, 2026); Reuters reporting by Anhata Rooprai and Jeffrey Dastin; TechCrunch; Engadget; 9to5Mac.