You can try Apple’s lightning-fast video captioning model right from your browser

Apple's FastVLM visual language model, now testable in-browser on Apple Silicon Macs, accurately captions live video in near real-time using a lightweight 0.5B parameter version. It runs locally, ensuring privacy for applications like wearables and assistive technology. FastVLM delivers video captioning up to 85 times faster than similar models while being three times smaller, demonstrating Apple's efficiency in AI optimization for their custom silicon.