Google published the July edition of Gemini Drops, its monthly roundup of Gemini app updates, on July 31[1]. The standout additions are an avatar feature that inserts your likeness into images without uploading a photo every time, and three new connected apps: Dropbox, Zillow Rentals and Viator. In the US, personalized image generation is now open to all users.
No more uploading a photo every time
The new avatar feature lets you add yourself to any image. According to Google, you can create content that looks like you without needing to upload your own picture on every attempt[1].
Until now, generating an image with yourself in it meant preparing and loading a source photo each time. Registering an avatar once removes that step. It should matter most for repetitive use cases, such as social posts or quick greeting cards where the same subject appears again and again.
Dropbox, Zillow Rentals and Viator join connected apps
Gemini has a framework called Connected Apps that calls external services with the user's permission. July's Drop adds three to that list: Dropbox, Zillow Rentals and Viator. Google frames it as linking your favorite apps to Gemini so a single prompt can carry several steps at once[1].
Zillow Rentals is the one with the clearest picture attached. On July 9, the US real estate platform Zillow announced that Zillow Rentals had gone live as a connected app inside Gemini and Gemini Spark, and that it is the only real estate platform in Gemini's connected apps ecosystem[2].
The flow starts as a conversation. Ask something like "Find me a pet-friendly one-bedroom in Seattle with tours on Zillow," and matching listings along with available booking times surface inside Gemini. One click sends the renter to Zillow to pick a time and lock it in[2]. What appears depends on the property's configuration: instant tour-enabled listings show a "Book a Tour" option, while the rest offer "Request a Tour." Renters connect their Google and Zillow accounts once, and the confirmed tour lands in their Zillow itinerary[2]. The experience is rolling out to Gemini and Gemini Spark users over the coming weeks[2].
The division of labor is clear: Gemini handles the conversation, Zillow handles what happens next. Michael Sherman, senior vice president of Zillow Rentals, said renters do not all start their search the same way, and what matters is that Zillow is present when they are ready to take a real next step[2]. Zillow reached ChatGPT in October 2025 and added conversational search to its own site five months later, so the Gemini integration sits on that same track[2].
Dropbox brings finding, summarizing, sharing and managing files and folders into Gemini, while Viator covers booking tours and activities. Neither is on by default. As connected apps, they only work once the user turns the connection on.
Personalized image generation reaches all US users
July's Drop also notes that image generation tailored to a user's own interests is now available to all users in the US[1].
The feature pairs Gemini's Nano Banana image model with Personal Intelligence, which handles user context. Because Gemini infers preferences from data tied to your Google account — Gmail, Google Photos, YouTube and Search history — the results reflect your tastes without spelling them out in the prompt[3]. Google's example: instead of "Create an illustration of me and my favorite things, such as coffee and baking," you can simply ask for "an illustration of me and my favorite things"[3]. Gemini can also pull actual photos of you from Google Photos[3].
The capability was previously limited to paid Plus, Pro and Ultra subscribers, and became free for all eligible US users on June 29[3]. Personal Intelligence is opt-in, and users choose which apps Gemini may access. Once enabled it becomes the default for every prompt, though a toggle in the Tools menu switches it off[3]. Better output comes with a wider range of personal data being read, which is worth keeping in mind.
Several items in July's Drop had already been announced individually. One is the expanded reach of Gemini Spark, with Google saying the agent that keeps working 24/7 even after you close your laptop is now available worldwide[1].
That "worldwide" has come with conditions, however. As of July 24, Google AI Pro subscribers had access in the US only, while Google AI Ultra subscribers had access globally with local language support, and the EEA, the UK, Switzerland and Nigeria were excluded[4]. Spark has been opening up plan by plan and region by region, so checking against your own subscription and location is the safer approach.
The remaining two items are voice control in Gemini for macOS and an update to the Flash model line. The former lets you speak into any active window to dictate text, rewrite highlighted content, or generate visuals right at the cursor[1]. The latter covers Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, which offer better reasoning and greater speed[1].
Summary
July's Gemini Drop leans less on new models and more on removing one step from routine work. The avatar removes a photo upload; connected apps remove the back-and-forth between multiple sites. At the same time, personalized image generation and the Zillow integration both start in the US, so not everything is available from Japan yet. Google keeps stacking these smaller changes through its monthly Drops, and the question for the next one is how much of this reaches other markets.
出典[1] https://blog.google/products-and-platforms/products/gemini/gemini-drop-july-2026/
出典[2] https://www.zillow.com/news/google-gemini-now-connects-renters-to-zillow-tour-booking/
出典[4] https://www.androidauthority.com/google-gemini-spark-wider-rollout-3690952/
