Google Releases 3 New Gemini Models, 3.5 Pro Still Not Available

Google quietly released three new Gemini AI models while leaving the eagerly awaited Gemini 3.5 Pro on the sidelines. The tech giant unveiled these models, which are designed to improve token efficiency and performance, signaling a strategic shift toward more resource-efficient AI as competition heats up with OpenAI and Anthropic. While the exact model names have not been officially disclosed, industry insiders confirm they span the nano, pro, and ultra tiers. The announcement came as a surprise to many, especially given the sustained anticipation for Gemini 3.5 Pro, which was expected to deliver significant reasoning improvements. These new models prioritize reducing computational costs while maintaining or even boosting output quality, a move that aligns with broader industry trends toward sustainable AI. Google has not provided a timeline for Gemini 3.5 Pro, leaving developers and enterprises in a holding pattern. The three new models are now available through Google AI Studio and Vertex AI, with pricing expected to be competitive against similar offerings from competitors. Early benchmark results suggest a 15–20% reduction in token usage for common tasks like summarization and code generation, without sacrificing accuracy. However, for complex reasoning chains, the models still lag behind the performance previews of Gemini 3.5 Pro. The release underscores Google's commitment to incremental innovation but raises questions about its roadmap for next-generation models. Analysts view this as a tactical move to address immediate market needs for cost-efficient AI deployment while continuing development on larger architectures. Looking ahead, all eyes remain on Gemini 3.5 Pro, which could redefine Google's competitive positioning. Enterprises should evaluate these new models for specific use cases where token efficiency is critical, while planning for the eventual upgrade to 3.5 Pro.