HomeWinBuzzer NewsMicrosoft Expands Azure OpenAI Capabilities with GPT-4 Turbo with Vision Public Preview

Microsoft Expands Azure OpenAI Capabilities with GPT-4 Turbo with Vision Public Preview

Microsoft's Azure OpenAI Service gets a vision upgrade. GPT-4 Turbo, once a text-only marvel, now analyzes images with GPT-4 Turbo with Vision.


OpenAI's GPT-4, the newest version of the large model, is now being enhanced by for its Azure OpenAI Service. Turbo with Vision has been made available as a public preview for Azure customers. The advanced AI model not only maintains the impressive language capabilities of its predecessor but also introduces the processing and analysis of image input. Microsoft has been planning the integration since last month

Expanding the Horizon of Azure AI

Microsoft has furnished its newly previewed multimodal AI model with several new features. One prominent feature is Optical Character Recognition (OCR), which enables the model to scan images, recognize, and extract text, integrating it seamlessly into the user's text prompts. This enhances the utility for a wide array of applications, from data interpretation to accessibility tools.

Another added feature is object grounding, allowing the AI to identify and reveal key objects within an image or specific video frames in response to textual inquiries. These developments in with Vision are designed to broaden the range of tasks the model can perform, creating opportunities for more sophisticated interactions between visual content and AI.

Improved User Experience and Accessibility

Utilizing the combined powers of GPT-4 Turbo with Vision, Azure AI Search, and Azure AI Vision, users can now incorporate images alongside text data. This enables the creation of solutions that enhance the chat experience by employing vector search to provide a connection with users' data. Microsoft assures that users engaging with the public preview of GPT-4 Turbo with Vision will receive an update to a stable, production-ready release in the forthcoming weeks.

As for the cost, the service is competitively priced at $0.01 per 1,000 tokens for Input processing and $0.03 per 1,000 tokens for Output processing. Prices are subject to change for the introduction of enhanced features. Presently, the GPT-4 Turbo with Vision's advanced features are accessible through certain regional data centers, including Australia East, Sweden Central, Switzerland North, and West US.

The advent of GPT-4 Turbo with Vision indicates a significant move by Microsoft to stay at the forefront of providing AI-powered solutions to a diverse customer base. This marks another milestone in the rapidly evolving domain of artificial intelligence, indicating the continuous push towards more seamless and complex integrations between various forms of data and machine learning capabilities.

Luke Jones
Luke Jones
Luke has been writing about all things tech for more than five years. He is following Microsoft closely to bring you the latest news about Windows, Office, Azure, Skype, HoloLens and all the rest of their products.