In the ever-evolving landscape of artificial intelligence, Google has introduced Gemini, a powerful tool that developers can now access through AI Studio (formerly MakerSuite). This article will guide you through the steps of using Gemini Pro via Google AI Studio, exploring its features, capabilities, and the potential it holds for developers.

Google AI Studio

Google AI Studio has undergone significant updates, now accommodating the Gemini Pro and Gemini Pro Vision models. This web-based tool offers developers the ability to generate both images and text, making it a versatile asset for various applications.

Quick Development

The tool is fully managed, providing a seamless programming experience for enterprises. A demonstration showcased a real estate broker application in AI Studio, utilizing natural language prompting with the added capability to upload images.

User-Friendly Interface

Accessing Google AI Studio is straightforward. Log in with your Google account and import files from Google Drive. Additionally, seamlessly transfer your existing code to your Integrated Development Environment (IDE) in supported languages such as Swift, Android (Kotlin), Python, JavaScript, and cURL.

Gemini Pro and Gemini Pro Vision

Developers can leverage Gemini Pro and Gemini Pro Vision for free, with a limitation of 60 requests per second. Google claims a 20 times higher appeal than competitors, coupled with cost-effectiveness due to upgraded Tensor Processing Units (TPUs).

Gemini V1.0

The initial version of Gemini Pro boasted a 32K context window. Google plans to enhance this window in future iterations. Trained on the Tensor Processing Unit, the AI hyper-computing infrastructure, Gemini V1.0 performs almost four times better than its predecessor, TPU v5p.

Features

Gemini V1.0 offers an array of features:

• Calling
• Embedding
• Semantic retrieval
• Custom knowledge grounding
• Chat functionality

SDK Support

Software Development Kits (SDKs) are available for Gemini Pro, supporting Python, Android (Kotlin), Node.js, Swift, and JavaScript. With support for over 38 languages across 1800+ countries and territories, Gemini V1.0 is accessible to a global developer community.

How to Use Google AI Studio

Getting started with Gemini Pro via Google AI Studio is a simple process. Follow these steps:

1. Visit ai.google.dev.
2. Click on “Get API Key” in Google AI Studio.
3. Sign in with your personal Google account.
4. Click on “Get API” in Google AI Studio, then click on “Create New.”
5. Choose “Freeform prompt” and start writing your prompt. Save it to generate a response using the model.
6. Change the Gemini model by clicking on the “Model” dropdown, selecting either Gemini Pro or Gemini Pro Vision.
7. To start using it, type at the prompt. For inserting an image, click “Insert” to upload the image.
8. Obtain the code for the generated content by clicking on “Get Code.” The API key can be acquired by clicking on “API key” in the left sidebar.

Pricing and Usage

While using Google AI Studio, developers can make up to 60 requests per second for free. Beyond this limit, a subscription to the paid tier is required. Pricing starts at $0.025 per 1,000 characters or $0.0025 per image. The output incurs a cost of $0.0005 per 1K character. Google AI Studio and Vertex AI offer connectivity to various corporate data sources.

Conclusion

As of now, specific pricing details for Gemini Pro are not available. Google plans to announce the pricing next year during the launch of Gemini Ultra. Additionally, Google is set to integrate Gemini into other developer tools like Android Studio, Colab, Firebase, and Flutter.

On another front, Google has upgraded its Google Cloud Vertex AI for enhanced control and integration. The Vertex AI Model provides over 130 AI models, including Imagen 2, ImagenMedLM, Gemini Pro, and Ultra, with more to be added in subsequent announcements.

FAQs

1. Is Gemini Pro available for free?
   • Yes, developers can use Gemini Pro for free with a limitation of 60 requests per second.
2. What languages does Gemini Pro support?
   • Gemini Pro supports Python, Android (Kotlin), Node.js, Swift, and JavaScript.
3. How does pricing for Gemini Pro compare to its competitors?
   • Google claims that their offering is 20 times more appealing than competitors, with cost-effectiveness due to upgraded TPUs.
4. Can I use Gemini Pro with other Google developer tools?
   • Yes, Gemini is planned to be integrated into other tools like Android Studio, Colab, Firebase, and Flutter.
5. What does the future hold for Gemini Ultra?
   • Specific pricing details for Gemini Ultra will be announced by Google next year during its launch.