Launched by Elon Musk in July 2023, xAI introduced its new multimodal model called Grok-1.5 Vision (Grok-1.5V), which represents a major step forward in the development of truly intelligent systems. announced. The company says the Grok-1.5V goes beyond text comprehension and integrates visual capabilities, allowing it to understand documents, diagrams, charts, photos and more.
What Does Grok-1.5 Vision Offer?
According to xAI, Grok-1.5V outperforms leading competitors in key benchmarks. Accordingly, the model performed excellently on xAI’s RealWorldQA dataset, an evaluation of more than 700 real-world images paired with questions. This dataset evaluates AI’s ability to understand complex visual scenarios and measures progress towards general intelligence.
The versatile perception of Grok-1.5V was demonstrated with examples such as turning a child’s drawing into a sleep story. The model also explained internet memes, converted charts to CSV format, and diagnosed problems with wood flooring based on images alone. xAI believes such a variety of tasks demonstrate the potential of the Grok-1.5V for a wide range of applications.
In the coming months, the company plans to focus its research on several key areas. xAI will soon call on the Grok-1.5V’s first testers to provide feedback and help improve its multimodal reasoning. Access to the Grok beta will initially be limited to X’s Premium+ subscribers, who receive additional benefits and support.
