Gemini 2.5 Model Expansion
Google has significantly expanded its Gemini 2.5 model family with the general availability of Gemini 2.5 Flash and Pro, marking their stable release for production applications. Simultaneously, the company introduced a preview of Gemini 2.5 Flash-Lite, positioned as its most cost-efficient and fastest 2.5 model to date. This new Flash-Lite version boasts enhanced quality across coding, math, science, reasoning, and multimodal benchmarks compared to previous iterations, offering lower latency for high-volume, latency-sensitive tasks like translation and classification. All Gemini 2.5 models maintain core capabilities including customizable “thinking” budgets, integration with tools such as Google Search and code execution, multimodal input, and a robust 1 million-token context length. These models are now accessible through Google AI Studio and Vertex AI, with Flash and Pro also available in the Gemini app, and custom Flash-Lite and Flash versions integrated into Google Search.
https://deepmind.google/discover/blog/were-expanding-our-gemini-25-family-of-models/