Anthropic folds chat and Cowork into one app
Anthropic's unified Claude interface auto-routes requests across chat, Cowork, Artifacts, presentations, and Docs, starting with Pro and Max plans.
Choosing between Claude's chat window and its Cowork workspace is about to stop being a decision users have to make. Anthropic is merging the two front-ends into a single interface, so that a request lands wherever the app decides it belongs rather than wherever the user guessed. The company says the change is aimed squarely at a specific friction point: people who opened the wrong tab and couldn't easily tell which one they were supposed to be in.
The rollout also brings dedicated tools for slide decks and documents, features Anthropic is positioning as part of the same push to consolidate scattered capabilities under one roof. Together, the changes describe an assistant that is being reorganized around what a user is trying to produce, not around which internal product surface they happen to be looking at.
A single window for everything
Before this rollout, Anthropic said customers often struggled to choose the right tab for the right task. The unified interface puts chat, Cowork, and Artifacts — Claude's interactive workspace feature — inside one window. Claude Design, introduced in April for website and prototype design work, also operates anywhere within Claude rather than being confined to a particular mode.
The mechanics matter more than the packaging. Instead of asking users to switch tabs or open separate windows, Claude now routes requests automatically. The stated goal is to remove the confusion around which feature to pick, using one front-end to dispatch work to the different parts of the app that handle it.
That is a meaningful shift in how the product presents itself. A routing layer that decides where a request goes is only as good as its judgment, and Anthropic has not detailed how the automatic routing makes its choices. What the company has described is the intent: fewer decisions for the user, less time spent figuring out the interface, more time spent on the actual task.
Slides and documents get dedicated tools
Anthropic is adding features built specifically for presentations and documents. Users can ask the assistant to create, edit, and present slides, and can download a generated presentation as a PDF or PowerPoint file. That export step gives the output somewhere to go beyond the chat window — a deck produced inside Claude can leave it in a format most workplaces already accept.
The Docs feature takes a different approach to the same problem. Rather than generating a finished artifact in one pass, it lets users work alongside Claude to create sections, ask questions, and leave comments on completed parts. It reads less like a generation button and more like a shared editing surface.
Both features support sharing and editing on a phone. Users can share generated documents or slides with a link, and they can start a document on desktop and then check the progress of the task from the mobile app. The handoff between devices is presented as a continuity feature rather than a sync feature — the same task, watched from a different screen.
Who gets it, and when
These features are rolling out to Claude's Pro and Max plans first, across web, desktop, and mobile, over the coming weeks. Anthropic plans to extend them to the free and team tiers later. The staged rollout is the familiar pattern: paying subscribers absorb the first version, and the broader base gets a more settled one.
That sequencing also means the unified interface will not arrive everywhere at once. Users on lower tiers will continue to see the older, split arrangement in the meantime, which could make comparisons between what different plans offer more visible than usual during the transition. Anthropic has not given a date for when the free and team tiers will follow.
An upgrade that landed just before
The interface change follows an upgrade to Cowork's memory layer, which helps Claude remember user context across chats. Memory and interface consolidation are connected problems. A workspace that forgets what you told it in a previous session forces users to compensate, and a split interface forces them to compensate again by tracking which tab holds which context.
By improving memory and collapsing the front-end in quick succession, Anthropic is addressing both halves of that burden. The upgrade of Cowork's memory layer was covered separately, and this rollout lands in its wake.
Neither change is presented as a new capability in the sense of a model that can do something it could not do before. Both are about reducing the overhead of using what already exists.
Why the tab problem was real
The company's own framing is that customers struggled to match tasks to tabs. That is a design failure more than a user failure. When a product grows by adding surfaces — a chat pane here, a workspace there, an interactive artifacts view somewhere else — the boundaries between them become less obvious to anyone who did not watch them being built.
Automatic routing is one answer to that. It accepts that users cannot be expected to maintain a mental map of the product's internal architecture, and moves the decision into the software. The trade-off is that the user gives up direct control over where a request lands, in exchange for not having to think about it.
Anthropic has not said whether users can override the automatic routing when they want a request handled somewhere specific. For now, the described behavior is that Claude routes requests on its own.
The presentation layer as the differentiator
Slides and docs are where the new features get concrete. Generating a deck is a common request, and the ability to export to PDF or PowerPoint determines whether the output is useful outside the app. Editing sections and commenting on completed parts, meanwhile, treats a document as something assembled over time rather than emitted whole.
Starting a document on desktop and monitoring it from a phone is a small detail that says something about the intended workflow: long-running tasks that a user can step away from and still supervise. Whether that matches how people actually work will show up in usage rather than in the announcement.
The features are aimed at the same paid tiers as the unified interface, which suggests Anthropic sees document and presentation work as a reason to subscribe rather than a baseline expectation.
What the consolidation signals
There is a broader read here about how AI assistants are maturing. The first phase of these products was about adding capabilities; the current phase, at least as Anthropic is playing it, is about making the existing ones findable. A unified front-end with automatic routing is less a new product than an admission that the previous one had accumulated too many doors.
The consolidation also concentrates risk in the routing layer. If Claude sends a request to the wrong part of the app, the user may not immediately know why the result looks different from what they expected. Anthropic has not described any visible indicator of where a request was routed, which means the routing may be invisible by design.
For competitors, the move is a reminder that interface complexity is a competitive surface. Anthropic's own framing invites the comparison, arguing that a single front-end is the better arrangement.
What this means for users and buyers
For individual subscribers on Pro and Max, the practical effect is fewer decisions and a shorter path from intent to output — assuming the automatic routing works as described. The document and slide tools add export formats that make Claude's output portable into ordinary work settings, which matters for anyone whose deliverables have to survive outside the app they were made in.
For teams evaluating AI assistants, the message is that feature counts are becoming less useful as a comparison metric than how much of the product a user has to understand before getting work done. Anthropic is betting that consolidation wins on that measure. Whether it does is an empirical question the company has not answered with numbers.
For everyone else, the changes are gated. Free and team tiers will keep the old arrangement until Anthropic extends the rollout, and no date has been given for that. The immediate audience is the paying one. This suggests the unified interface will be tested under real usage before it reaches the wider base — and that any rough edges in the routing logic will surface first among the people most invested in the product.
Sources
- TechCrunch Original source
- upgrade of Cowork’s memory layer Also reporting
Continue Reading
Rift Reshapes Enterprise AI Access
A public split among AI labs over safety testing may alter how enterprises access and deploy frontier models, analysts say.
Amazon Rolls Out Alexa+ Across India
Amazon has launched its Alexa+ assistant in India with Hindi and English support, priced at ₹2,000 a month for non-Prime users.
AI in cybersecurity: jobs shift, SOCs change
Experts say AI will transform security operations, vulnerability triage, and roles, but reproducibility and governance remain hurdles.