Google's VP of Gemini, Josh Woodward, took the unusual step of asking users to publicly list what frustrates them most about the app, then published a ranked breakdown of the top 10 complaints and Google's plans to fix them. More than 1,400 users responded within the first 12 hours, and Woodward says he read every single one before committing to specific improvements.

What Are Gemini Users Most Frustrated About?

The number-one complaint was that Google Workspace integrations, which include Gmail, Docs, Slides, Keep, and Tasks, don't work reliably enough. This stings for Google because the company heavily markets Gemini as an assistant that seamlessly moves across all these tools. Woodward acknowledged the problem directly and noted that improvements have recently landed in the Gemini Spark agent, though reliability across the whole app still needs significant work.

The second and third most common complaints centered on more reliable tool calling and better project and folder organization for chats. Woodward said Google "strongly" agrees with the tool-calling feedback and promised noticeable improvements soon, likely tied to future model upgrades. For chat organization, Google acknowledged that its Notebooks feature isn't enough and committed to rethinking the entire experience.

How Is Google Addressing These Issues?

• Workspace Integration Reliability: Google has deployed recent improvements in Gemini Spark but admits the whole app needs far better reliability across Gmail, Docs, Slides, Keep, and Tasks.
• Tool Calling Improvements: Google plans noticeable gains soon, with enhancements likely riding on future model upgrades that will improve how Gemini connects to external services.
• Chat Organization: Google will rethink the entire project and folder experience beyond the current Notebooks feature and report back with a better solution.
• Message Editing: Users requested the ability to edit messages after sending them, and Google confirmed this feature is in progress.
• Voice Dictation: Sharper voice dictation was flagged as a pain point, and Google is working on improvements.
• Mobile Scrolling Bugs: Fixes to mobile scrolling issues are confirmed to be in progress.
• Model Connections (MCPs): Google has already rolled out early support in Gemini Spark with third-party connections including Canva, Dropbox, Instacart, OpenTable, and Zillow.

Beyond these seven items, users also requested the ability to export Deep Research reports to NotebookLM and swap between models mid-chat. Woodward admitted he hadn't even known these were pain points until users raised them.

What Is Google Refusing to Change?

Not every request made the cut. Two items sparked particular attention. First, users wanted watermarks removed from images generated by Nano Banana, Google's most cost-efficient image generation model. Woodward said Google would weigh this against AI safety rules that differ by country, leaving the decision uncertain.

The one flat refusal came on celebrity likeness guardrails. Google plans to keep those protections exactly as they are, so users hoping to create deepfakes or celebrity-lookalike content will need to look elsewhere.

What New Gemini Models Did Google Just Release?

While addressing user complaints, Google also launched two new AI models on June 30 designed to expand its image and video capabilities. Nano Banana 2 Lite is positioned as the company's most cost-efficient image generation model, capable of generating images from text in about four seconds at a price of $0.034 per 1,000 images. The model is available through developer platforms, AI mode in Search, NotebookLM, Google Flow, Google Ads, the Gemini app, Google Photos, and Stitch.

The second model, Gemini Omni Flash, is built for high-quality video generation and conversational video editing. It is now available to developers through the Google AI Studio and Gemini API, priced at $0.10 per second of video output, matching the cost of competing video models.

"With these two models, developers can build comprehensive, end-to-end multimedia experiences that connect rapid image generation with video creation and editing," Google stated in a release.

Google, Official Release

The timing of Woodward's public accountability push comes after May's Neural Expressive redesign, suggesting Google is listening more closely to user feedback than usual. This rare bit of transparency from a company that typically prefers to control its own messaging signals that Google recognizes the competitive pressure in the AI assistant space and is willing to engage directly with its user base to improve the product.