Elon Musk's xAI releases Grok AI with vision and multilingual capabilities

Grok Vision allows real-time interaction through a device's camera; multilingual and voice search features introduced for paid users

Elon Musk’s artificial intelligence company xAI has introduced a new feature for its Grok chatbot called Vision. This feature allows the chatbot to analyse real-world objects, text, and environments through a user’s device camera in real time. Available for free on Android and iOS, Grok Vision offers functionality similar to features already available in OpenAI’s ChatGPT and Google Gemini.

 

According to xAI, Grok Vision enables users to scan and search anything in real time—including objects, symbols, and documents—by using their device’s camera and asking questions about them. For example, users can show Grok a business card, office gate, or watch to receive instant information. As the camera moves, Grok responds with relevant answers in real time.

 

 

Grok officially announced the update on X (formerly Twitter), stating: “I am all ears (and eyes). Share your world with me.”

 

i'm all ears (and eyes)  share your world with me https://t.co/UuKhy54asa

— Grok (@grok) April 22, 2025

New features: Multilingual audio and voice search

 

Alongside Grok Vision, xAI also launched two additional features: Multilingual audio and real-time search in Voice Mode. The multilingual feature allows Grok to respond in several languages, including Hindi, Spanish, and Japanese. The voice search feature lets users ask questions aloud, with Grok retrieving fresh information and reading it back to the user.

 

These features are currently available to paid subscribers of the SuperGrok plan, which costs $30 per month. It is currently unclear whether these features will be extended to non-subscribers.

Memory function for personalised responses

 

In related developments, Grok recently introduced a memory function that enables it to recall earlier interactions, offering more personalised and context-aware responses. According to xAI, this feature improves Grok’s ability to tailor its replies based on a clearer understanding of individual user preferences developed over time.