Grok can now generate videos with audio: Here's how
What's the story
Elon Musk has demonstrated the latest capabilities of his artificial intelligence (AI) platform, Grok. He shared a whimsical prompt on X, reading: "A park ranger taking a photo of a family of four adults and children dressed in shorts and t-shirts posing by their camper van in a national park, with a smiling sasquatch standing in the woods. Everyone is eating bananas." The post was accompanied by an equally quirky video generated from this prompt, showcasing Grok's creative potential.
Twitter Post
Take a look at Musk's post
Grok Imagine prompt:
— Elon Musk (@elonmusk) September 7, 2025
A park ranger taking a photo of a family of four adults and children dressed in shorts and t-shirts posing by their camper van in a national park, with a smiling sasquatch standing in the woods.
With added speech prompt:
“Everyone is eating bananas” pic.twitter.com/HMy50oufiv
Platform evolution
Grok can now generate talking videos
Last week, Musk had teased a major upgrade to Grok AI, saying "Grok videos can now talk." The update marks an expansion of both image and video generation capabilities. Musk described these enhancements as an "early beta," suggesting that more improvements are on the way in the coming weeks. The demonstration from DogeDesigner account on X showed an anime character introducing itself as Anne, highlighting the platform's transition from static image generation to interactive content.
Tool enhancement
Grok AI's latest upgrades and prediction capabilities
Grok AI has been gaining traction for its predictive and creative capabilities. The latest update makes the platform an even more versatile tool for content creators, enabling them to produce more imaginative and interactive outputs. Along with video/audio generation, Grok is set to be a major player in the AI-driven content creation space. Musk also highlighted Grok's forecasting prowess by sharing a link to a live benchmark platform on his official X account, encouraging users to test its prediction abilities.