Deep Research Feature Removed from ChatGPT Desktop App Amid Feature Overhaul

OpenAI has removed the Deep Research feature from the ChatGPT desktop application, a change confirmed by users on Reddit and observed across multiple platforms as of February 2026. The feature, previously available as a premium research tool enabling deep web analysis and citation-backed responses, is no longer visible in the desktop app’s interface, though it remains accessible via the web version. The removal has prompted widespread discussion among power users, researchers, and enterprise clients who relied on the functionality for academic, journalistic, and market analysis tasks.

While OpenAI has not issued an official statement regarding the removal, its latest release notes—published on its Help Center—highlight significant updates to context window capacity and interactive code blocks, suggesting a strategic pivot toward enhancing core conversational and coding capabilities over specialized research modules. According to OpenAI’s official changelog, the ChatGPT model now supports a 256k token context window (128k input, 128k output) when users activate the ‘Thinking’ mode, up from the previous 196k. This expansion allows for more comprehensive document analysis and longer conversational threads, potentially reducing the perceived need for a separate research module.

Meanwhile, the integration of interactive code blocks—allowing users to write, edit, and preview code and visualizations directly within chat—indicates OpenAI’s increasing emphasis on productivity and real-time collaboration. These enhancements align with broader industry trends toward AI tools that serve as integrated workspaces rather than standalone research assistants. The removal of Deep Research may reflect an internal assessment that its functions are being absorbed into the model’s enhanced reasoning and retrieval capabilities.

Notably, the feature’s disappearance coincides with growing competition in the AI research space. While DeepL continues to expand its AI translation and agent tools, and media outlets like MSN highlight prompt-based research techniques that leverage ChatGPT’s existing capabilities, OpenAI appears to be consolidating its feature set. A recent MSN article, for instance, details how a single well-crafted prompt can now replicate hours of market research using ChatGPT’s built-in web browsing and data synthesis—techniques that may have rendered the dedicated Deep Research interface redundant for many users.

Users who relied on Deep Research for structured citations, source verification, or multi-source synthesis are now adapting. Some have turned to third-party plugins or manual prompt engineering to replicate the feature’s functionality. Others have expressed concern over transparency, noting that OpenAI’s lack of public communication on the change undermines trust in its product roadmap. The absence of an announcement or migration guide has left many feeling blindsided, particularly in professional settings where workflow continuity is critical.

Industry analysts suggest the move may be part of a larger effort to streamline ChatGPT’s interface for broader consumer adoption. By reducing feature bloat and focusing on core competencies—contextual reasoning, code generation, and real-time interaction—OpenAI may be preparing for a future where AI assistants are embedded into everyday productivity suites rather than treated as specialized tools. The future of AI-powered research may lie not in standalone features, but in the model’s ability to intuitively understand and execute complex tasks through natural language.

As of now, OpenAI has not indicated whether Deep Research will return in a reimagined form or be replaced by a new architecture. Users are encouraged to monitor official release notes and consider using the web version of ChatGPT, where the feature remains active—for the time being.