Google Drops 3 New Gemini Models, Skips 3.5 Pro

Google has quietly released three new Gemini models, expanding its AI portfolio without the much-anticipated Gemini 3.5 Pro. The new additions—Gemini 2.0 Flash, Gemini 2.0 Pro, and Gemini 2.0 Ultra—arrive with distinct capabilities tailored for different use cases. Flash is optimized for speed and cost-efficiency, Pro balances performance and scalability, while Ultra targets complex reasoning tasks. Notably, the absence of a 3.5 Pro update signals a strategic pivot, possibly toward incremental improvements rather than a major version leap.

This release pattern suggests Google is prioritizing specialization over a single flagship model. By offering tiered options, the company aims to capture a broader developer and enterprise audience, from lightweight applications to high-end research. The move also aligns with the industry trend toward modular AI systems, where users can choose the right model for their specific needs—whether it's real-time video generation, text analysis, or multimodal tasks.

What This Means for AI Video Creation

For the AI-video space, the implications are significant. The new Gemini models are expected to enhance video understanding and generation capabilities, though specifics remain under wraps. The absence of a 3.5 Pro model suggests Google is focusing on refining its existing architecture rather than rushing a major update. This could mean more stable, reliable performance for video tasks like scene segmentation, captioning, and style transfer—critical for creators using EZ Magic Video. However, without a flagship model, competitors like OpenAI's Sora or Meta's Make-A-Video might gain an edge in raw innovation.

The strategic takeaway is clear: Google is betting on incremental, specialized improvements over a single headline-grabbing release. For developers and content creators, this means a more predictable, less disruptive upgrade cycle—but also potentially slower access to cutting-edge features. The lack of a 3.5 Pro model might disappoint some, but it signals that Google is prioritizing stability and real-world performance over benchmark scores. As the AI video landscape evolves, these models could provide the reliable foundation needed for production workflows, even if they don't steal headlines.