Claude Cowork Is Gone: Why 'One Claude' Should Have Happened Months Ago

Anthropic just fixed the most confusing thing about its own product line: as of September 16, Claude Cowork and Claude chat are merging into one Claude — 'Bring a quick question, or hand over a report due at noon, and Claude takes it from there, even after you've closed your laptop,' in the company's words — alongside three new betas on paid plans: Claude Docs, Claude Slides, and Claude Design working directly inside conversations.

Our take up front, because we've held it for months: this merge is overdue, and the fact that it happened is Anthropic admitting the split never made sense. (Disclosure, per The Receipts Standard: this site is produced with Claude-family tools — which is partly why we're confident about the confusion; we lived in it.)

What actually changed — and what's permanent

The mechanics: one conversation surface, with Claude itself deciding whether your request is a quick answer or a longer background job — 'more involved tasks keep running in the cloud even if you close your laptop,' per the help center, with Claude asking a clarifying question when scope is ambiguous, and multiple tasks running in parallel. Existing Cowork tasks, projects, connectors, skills, and files carry over. Rollout: Pro and Max first over the coming weeks, Team and Free later, and Enterprise admins get at least 30 days' notice. Claude Code stays separate.

The new products: Claude Docs is a real collaborative document editor — live co-editing, comments with @Claude mentions, export to Word, Google Docs, PDF, or Markdown. Claude Slides drafts presentations you can edit, present from Claude, or download as PowerPoint or PDF (note: no Google Slides export). And Claude Design now runs in-chat, with the standalone version continuing.

Two things are worth reading twice: the merge is one-way — 'once your account has the new experience, you can't switch back' — and everything, quick or long, 'counts toward your plan's usage limits,' with Anthropic noting usage 'may be measured slightly differently' during the rollout. More on that below.

Our take: this should have happened months ago

When Cowork launched in January as 'Claude Code for the rest of your work' — a separate place you'd deliberately switch to for 'bigger work' — the theory was that people would pre-sort their tasks. They didn't, because nobody thinks that way. No knowledge worker looks at a request and classifies it as 'chat-sized' versus 'Cowork-sized' before starting; you find out how big a task is by doing it.

The distinction was product architecture wearing a user-benefit costume, and Anthropic's own launch post now concedes it: people 'told us the frustrating part was deciding where a task belonged.' The tell that even Anthropic couldn't keep the story straight: Enterprise accounts had Cowork switching on by default starting September 10 — six days before it stopped being a separate product.

The auto-routing model is the right abstraction (it's how Claude Code has always worked: you ask, it decides how hard to work), and the eight months of 'which Claude do I open?' were a tax on exactly the mainstream users Anthropic needs. Good merge. Late merge.

The honest catches

Three, none fatal, all worth knowing. One: no exit. The merge is mandatory once it reaches your account — fine for most, worth flagging for teams with Cowork-specific workflows.

Two: the meter got fuzzier. When the model decides whether your question becomes a multi-step cloud task, the model is also deciding how much of your usage limit it spends — 'longer agentic tasks... generally use more than a quick question,' as the help center puts it, and Fortune's launch-day coverage flagged the same concern: agentic work burns tokens. We'd like to see clear in-product cost signals before a task runs long; watch this space on the Pro tier especially.

Three: the betas are betas. Claude Docs currently has no version history, no comment-only permissions, no external sharing on Team/Enterprise, and isn't available under HIPAA-ready or customer-managed-key configurations — real limits for exactly the office workers this launch courts.

The Google collision — and the Gamma awkwardness

The Verge's headline is the market's framing — 'Claude comes for Gemini with its own take on Docs and Slides' — noting that Anthropic's models are widely seen as stronger while 'Google's AI services have the advantage of being natively tied into its suite,' and that Docs and Slides give users 'fewer reasons to leave a Claude chat.' That's the fight: model quality versus workspace gravity, and every 'everything app' move like this shrinks the reasons to round-trip to Google.

The subplot nobody's covering: Gamma is an Anthropic customer — its decks are literally built on Claude's API — and Anthropic just shipped a first-party competitor to it. That's the platform-company pattern (AWS ate its ecosystem the same way), and every AI-app builder should read it as the strategy it is.

As for whether Claude Slides is actually good: unconfirmed by anyone yet, including Anthropic's own launch post, which is conspicuously quiet on output quality.

Bottom line: the merge is the story for users — one Claude, no taxonomy homework, work that survives a closed laptop — and it's the right call arriving late. Docs, Slides, and Design are the story for the market: Anthropic is done being a chatbot with a great model and is building the office suite around it.

Where this touches our rankings: Claude sits #2 in our presentation makers ranking largely on its real-.pptx-everywhere story, and Gamma holds #1 on finished-deck quality. Claude Slides reaching our plans triggers a full re-test of that page — and 'Claude Docs vs Google Docs' just entered our comparisons pipeline. When the betas land, we'll publish what the decks actually look like, per the usual standard: receipts, not vibes.