Google is testing unreleased changes to close feature gaps between the Gemini desktop app and its web version, including dedicated tabs for image/video generation, camera capture, and custom MCP server support for Spark.
Google is enhancing the Gemini desktop app by adding dedicated tabs for image and video generation and introducing a camera attachment feature for capturing photos.
# Google is aiming to close feature gaps on Gemini desktop
Source: [https://www.testingcatalog.com/google-is-aiming-to-close-feature-gaps-on-gemini-desktop/](https://www.testingcatalog.com/google-is-aiming-to-close-feature-gaps-on-gemini-desktop/)
Google is narrowing the gap between the Gemini desktop app and its web counterpart, with several unreleased changes now in testing\. Dedicated tabs for**image and video generation**are being added to the app's navigation, and an app configuration panel has appeared on the desktop for the first time\.
A second track is a**camera**attachment\. From the attachment menu, users would select a camera option, which opens a capture screen, takes a picture, and drops it into the prompt\. It reads as photo capture rather than a live video stream, keeping it distinct from[Gemini](https://www.testingcatalog.com/tag/gemini/)Live's real\-time sharing and closer to what mobile users already have\.
The third piece sits with[Spark](https://www.testingcatalog.com/google-unveils-24-7-gemini-spark-ai-agent-for-advanced-tasks/)\. Its connectors menu on desktop now includes an option to add custom MCP servers, bringing something previously confined to internal configuration into the visible agent settings\. Google already permits custom MCP apps for Spark via the web app, so this is an extension of that rather than the first appearance, though it matters for anyone running Spark from the Mac app\.
> Gemini Spark is rolling out to Google AI Pro users outside the U\.S\. Spark is your personal AI agent that works in the background 24/7 to get things done under your direction, handling the heavy lifting so you can focus on what matters\.[https://t\.co/qq1RQK221m](https://t.co/qq1RQK221m?ref=testingcatalog.com) — Google Gemini \(@GeminiApp\)[July 31, 2026](https://x.com/GeminiApp/status/2083302569796059271?ref_src=twsrc%5Etfw&ref=testingcatalog.com)
Timing is unclear; these are trusted\-tester builds with no public rollout date\. The context is a busy month: Spark reached macOS on July 1 alongside Google Tasks and Keep connections, and intelligent dictation and screen\-aware reasoning began rolling out on July 29\. Together, generation tabs, camera capture, and connector management would leave fewer reasons to open the browser\.
Google is preparing to bring Gemini Live real-time voice mode and customizable Skills to the web version of Gemini, closing feature gaps with competitors like ChatGPT and Claude. Testing is underway, and a rollout could happen by late July 2026.
Google is developing a new 'App' feature for Gemini Notebook that turns source materials into interactive applications like dashboards or study aids, alongside other refinements such as AI Notes and watermarking.
Google announces major updates to its Gemini app at Google I/O, including a Daily Brief feature, redesigned Neural Expressive interface, a personal AI agent called Gemini Spark, and integration of the new Gemini Omni video model to compete with ChatGPT and Claude.
Google is preparing Plugins and a Notifications area for Gemini Enterprise, aiming to package reusable workflows with connectors and skills for business users. The features are still in development with no official announcement or rollout date yet.
Google has made Gemini 3.1 Flash-Lite generally available, offering ultra-low latency and high-volume processing with multimodal capabilities, targeting enterprise applications.