#Google Gemini - Page 2
45 Stories
Arjun Sha
Recently, Google released the API key for its Gemini Pro model. Currently, the company is offering text-only and text-and-vision models based on the Gemini Pro model. The best part is that Google is allowing users to test the API for free and without setting up Google Cloud billing, at least for now. To teach you how to set up the Gemini AI API key and use it, we have included three coding examples to showcase the API usage. You can test Gemini Pro's multimodal capability as well through the API, which is not yet available in Bard.
Google has finally released a true multimodal AI model called Gemini. While multimodal features are not live yet, you can use Bard to check out the Gemini Pro model. Pixel users can also experience on-device AI with the Gemini Nano update. The most powerful Gemini Ultra model is quite impressive and beats the GPT-4 model on several benchmark tests. In multimodal tests too, Gemini Ultra dethrones GPT-4V model. However, the Ultra model has not been launched yet and is due to go live early next year.
Extensions on Gemini allow you to connect to many Google services. In your conversations, you can chat with Gemini and pull data from YouTube, Google Flights, Gmail, Drive, Docs, and more. Keep in mind, your personal data from Gmail, Drive and Docs is not used for training Gemini. You can even use the YouTube extension to chat and summarize YouTube videos.