Google has announced that developers now have access to a 2 million context window for Gemini Pro 1.5, compared to GPT-4o’s 128k context window. This feature was initially available through a waitlist but is now accessible to everyone. To mitigate higher costs associated with longer context windows, Google introduced support for context caching in the Gemini API for Gemini 1.5 Pro and 1.5 Flash, allowing context to be stored for later use and reducing costs for tasks that reuse tokens across prompts.
In addition, Google enabled code execution for both Gemini 1.5 Pro and 1.5 Flash, enabling the model to generate and run Python code iteratively until the desired result is achieved. The execution sandbox is not connected to the internet, comes with pre-installed numerical libraries, and developers are billed based on the output tokens from the model.
Furthermore, Gemma 2 is now available in Google AI Studio, and Gemini 1.5 Flash tuning will be accessible via the Gemini API or Google AI Studio in the coming month. These advancements aim to enhance the capabilities and flexibility of Gemini models for developers.
Source link
Source link: https://sdtimes.com/ai/google-launches-2-million-context-window-for-gemini-1-5-pro/