Google is set to launch two new AI models: Gemini 3 and Nano Banano 2. Gemini 3, optimized for coding and regular tasks, is expected to roll out in November 2025, with early reports indicating a context limit of 1 million tokens. Although rumors suggest access to Gemini 3, Google typically keeps its testing discreet, allowing only select partners to test under strict NDAs.
Currently, the latest model is Gemini 2.5 Pro, which remains competitive in AI performance rankings, scoring 63.8% on the SWE-Bench Verified standard.
Additionally, Google is developing Nano Banano 2, codenamed “GEMPIX2,” aimed at generating highly realistic images, potentially launching by December 2025. It’s important to note that rival companies like OpenAI are also advancing their AI offerings, including GPT 5.1.
Stay updated with the latest AI trends and improvements to ensure your applications remain competitive and secure.
Source link
