Detailed Analysis
A Reddit user's query about running multiple Claude Desktop windows simultaneously surfaces a persistent gap in Anthropic's desktop application: the lack of native multi-window support for parallel workflows. The poster describes a common power-user pattern—juggling several concurrent chat sessions, "cowork" collaborations, and coding tasks—and wanting to view and manage them side by side rather than switching between tabs or sessions within a single window. Notably, they identify a partial workaround already present in the Code view, where a three-dot menu offers an "Open In" → "New Window" option, but this functionality is scoped narrowly to code sessions and doesn't extend to general chat or cowork contexts, leaving users without a consistent, application-wide solution.
The request highlights a functional limitation that becomes increasingly salient as Claude's use cases expand beyond single-threaded conversation into more complex, multi-context work. As Anthropic pushes Claude toward agentic coding, long-running tasks, and collaborative "cowork" features, users naturally begin working on several parallel threads—debugging one project while drafting content in another, or monitoring a long-running agent task while starting a new query. Desktop applications from competitors and general productivity tools have long supported multi-window or multi-instance operation precisely because knowledge workers rarely operate on a single task at a time. The absence of this in Claude Desktop suggests the product's architecture may still be optimized around a single-session mental model inherited from its browser-based origins, rather than a native desktop paradigm.
The user's observation that this feature "only benefits them" from Anthropic's perspective—since more concurrent sessions would likely translate to increased token consumption and usage—is a sharp bit of product commentary. It implicitly questions why a low-effort, revenue-positive feature request would go unaddressed, and raises the possibility that engineering resources are being prioritized elsewhere, such as model capability improvements, safety work, or higher-visibility features. This kind of gap between user expectations and product reality is common in fast-moving AI companies where the pace of model development often outstrips the maturation of surrounding tooling and UX polish.
More broadly, this thread reflects a recurring theme in the AI assistant space: as tools like Claude become embedded in daily professional workflows, users increasingly expect them to behave like full-fledged productivity software rather than simple chat interfaces. The demand for multi-window support, session management, and better workspace organization signals that Claude's user base is maturing past casual conversational use into sustained, multi-threaded professional work—coding, research, writing, and coordination happening in parallel. Anthropic's ability to close these workflow gaps, alongside its headline advances in model intelligence, will likely shape competitive differentiation against rivals like OpenAI's ChatGPT desktop app and other AI-native tools that have already invested more heavily in multi-session and window management capabilities.
Read original article →