xAI, Elon Musk's AI endeavour, has introduced Grok-1.5V, the company's first-gen multimodal model. With the enhancements, Grok will be able to process visual data such as photos, diagrams, charts and more. The update comes a few days after xAI announced Grok-1.5 with improved reasoning and context length. It will be available for X (formerly called Twitter) users soon.

According to xAI's blog, Grok 1.5 Vision Preview will soon be available for testers and users. The multimodal model can understand a range of imagery such as screenshots, images, and documents. Grok 1.5V outperforms the competition in the RealWorldQA benchmark that takes into account real-world spatial understanding, according to the AI business.

Also Read: Grok-1.5 AI Will Be Available Next Week, Grok 2 AI Now In Training, Confirms Elon Musk

Based on the examples shared in the blog, the new version can write code from a diagram, calculate calories from nutrition facts, frame stories from drawings, explain memes, convert tables in images to CSV files and more. "We are introducing a new benchmark, RealWorldQA. This benchmark is designed to evaluate basic real-world spatial understanding capabilities of multimodal models," the company said.

The initial release consists of 700+ image datasets from the real world. "We are excited to release RealWorldQA to the community, and we intend to expand it as our multimodal models improve," x AI said. Meanwhile, Grok 1.5 which comes with better reasoning and a context length of 1,28,000 tokens.

Also Read: Elon Musk-Led xAI Open Sources Its AI Chatbot Grok For Developers | All You Need To Know

xAI claims that Grok 1.5 offers significant improvements in coding and maths. It offers up to 16 times larger memory capacity than the previous version to digest more data. Grok 1.5 will roll out to early testers and Grok users on X soon. xAI said that the company will also roll out more features in the coming weeks. In this endeavour, the company is also hiring AI engineers, tutors, designers and more. Plus, Grok 2 is reportedly already in the works.


Also In News