Google Gemini | July Drop Expands Personalized Images
Google's July Gemini Drop expands visual creation with reusable avatars and image generation shaped around a user's interests and preferences. The update also introduces voice-based creation on macOS, new Flash models, worldwide access to Gemini Spark in supported regions, and additional app connections for completing tasks through a single prompt.
Gemini makes image generation more personal
Gemini can now add a user's avatar to generated images without requiring the same reference photo to be uploaded for every request. Once the avatar is available, creators can place themselves in different visual concepts while keeping the generation process inside the Gemini app.
This creates a more direct workflow for personalized social graphics, profile concepts, presentation visuals, and other content where the user needs to appear in the composition. The final result should still be reviewed for facial consistency, clothing details, framing, and whether the generated scene matches the intended identity.
Personal preferences can guide new visual concepts
Gemini can also generate images informed by a user's interests and preferences. Instead of treating every request as an isolated prompt, the app can use that personal context to prepare visuals that feel more closely connected to the subjects, aesthetics, or ideas the user regularly explores.
The feature was made available to all users in the United States as part of the July Drop. Creators outside that market should check their Gemini account before planning a workflow around it, since the announcement does not confirm the same availability for other regions.
Voice creation comes to active windows on macOS
Mac users can now speak to Gemini while working inside an active window. The feature can dictate cleaned-up text, transform highlighted content, summarize material, or generate visuals at the current cursor position, reducing the need to move between the open project and a separate AI conversation.
For creative work, this may help when drafting image directions, rewriting captions, condensing reference material, or requesting a visual while staying focused on the current document. Because spoken instructions can omit important details, prompts and generated results should still be checked before they are used in a finished design.
New Flash models support faster everyday work
The July update also brings Gemini 3.6 Flash and Gemini 3.5 Flash-Lite to the app. Google positions the new Flash models around improved reasoning and greater speed, giving users additional options for tasks that need a balance between response quality and quick iteration.
Gemini can now connect with Dropbox, Zillow Rentals, and Viator as well, allowing information and actions from supported services to be coordinated through one prompt. These connections are broader productivity additions, but Dropbox may be especially useful when creative references and working files are already stored there.
Gemini Spark expands with regional exceptions
Gemini Spark is expanding worldwide and can continue working on assigned tasks after the user closes their laptop. This gives creators a way to leave longer-running work active instead of keeping the original device open throughout the process.
The worldwide description has specific exceptions. Gemini Spark is not available in the European Economic Area, the United Kingdom, Switzerland, or Nigeria, so access should be confirmed before depending on it for unattended work.
IMPORTANT: Personalized image generation based on user interests was announced for all users in the United States. Gemini Spark excludes the European Economic Area, the United Kingdom, Switzerland, and Nigeria.{alertWarning}
Daisuki's Take: What This Means for Designers
The most relevant change for visual creators is the move from isolated image prompts toward reusable personal context. An avatar can reduce repetitive reference uploads, while preference-based generation may provide a closer starting point for recurring visual themes.
Personalization does not remove the need for art direction. A result can resemble the user or reflect familiar interests and still miss the intended composition, visual hierarchy, style, or level of consistency required for a real project.
We would use these features for exploration and early drafts, then review every visual as part of the complete design rather than judging personalization alone. Regional access, identity accuracy, privacy preferences, and the suitability of each generated image should be checked before making the workflow repetitive or public.
Sources and Recommended Links
- Gemini Drops: New Updates to the Gemini App, July 2026 | Google Official Blog
