Google has been steadily updating the Gemini app for Android with small improvements, but the experience still trails what the underlying model is capable of. The company appears to recognize that gap. In response to a user’s criticism on X about Gemini’s interface lagging behind ChatGPT, Logan Kilpatrick — who leads product efforts for Google AI Studio and the Gemini API — said Google is currently making a “huge investment” in Gemini App UX 2.0. It’s a rare, candid acknowledgment that the front end needs more than incremental polishing.
The current app hides many of Gemini’s stronger features behind menus and workflows that don’t always feel intuitive. A redesign could bring those capabilities forward, making the system easier to use for everyday tasks instead of something users have to dig for. Kilpatrick also mentioned that Google is working on a native Gemini app for macOS, which would give desktop users a more direct experience than the browser-based setup available today.
The comparison to ChatGPT is hard to avoid. OpenAI offers native apps for both Windows and macOS, which gives users faster file handling and tighter integration with system-level tools like Notion, Terminal, and Apple Notes. Gemini, by contrast, remains limited to the browser on desktop, which works but can feel slower and less integrated. Something as simple as uploading multiple files, for instance, becomes smoother on ChatGPT’s desktop apps, while Gemini users have to navigate browser windows and file pickers.
Native software also becomes more important as AI models begin moving toward agent-like behaviors that involve working with local files, automating tasks, or interacting with other apps. Browser access is functional, but native access is more flexible. Google’s effort to bring Gemini to macOS suggests the company understands this shift and wants the platform to keep pace. There’s no timeline yet, though Google’s recent update cadence hints that the wait may not be long.
Beyond Gemini itself, Google is also building a mobile app for AI Studio, the company’s environment for developing and testing models. Dubbed Build Anything, the app is in early stages according to Ammar Reshi, who leads product and design for Google AI Studio. A dedicated mobile version could help developers continue experimenting and coding while away from their laptops, though it doesn’t appear close to release.
What’s clear across these moves is that Google is treating Gemini’s interface — not just the model — as an area that needs substantial rethinking. Whether Gemini App UX 2.0 and a new round of native apps can make the platform feel more cohesive is a question that will likely be answered over the coming year.
