Google Unveils Next-Gen ‘Gemini Pro’ AI Model for Developers

Last week, Google announced its next-gen Gemini AI model in three sizes: Ultra, Pro, and Nano, and today, it is making the Pro version available for developers to build for their own use cases.

Google is already integrating the model into various products such as Android’s Pixel 8 Pro and the specially tailored Bard version of Gemini Pro.

Gemini Pro Features

Gemini Pro, designed to outperform other models of similar size, currently offers a 32K context window for text. Future updates promise an even broader context window and competitive pricing

This version boasts diverse features like function calling, embeddings, semantic retrieval, and chat functionality, supporting 38 languages across 180+ countries.

Developers can now leverage Gemini Pro through the Gemini API, accepting text inputs and generating text outputs.

A dedicated Gemini Pro Vision multimodal endpoint is also available, receiving text and imagery as input with text output.

Google AI Studio

Additionally, Google AI Studio, a free web-based developer tool, facilitates prompt creation and API key acquisition for app development.

It offers a generous free quota of 60 requests per minute, twenty times more than other free offerings.

Upon completion, users can transfer their work to their preferred Integrated Development Environment (IDE) or use quickstart templates in Android Studio, Colab, or Project IDX.

Gemini Pro Pricing and Access

Currently, developers enjoy free access to Gemini Pro and Gemini Pro Vision via Google AI Studio.

Vertex AI developers can avail themselves of the same models at no cost until early next year, after which charges per 1,000 characters or images will apply across both platforms.