Anthropic models do not support the video modality as input, but through skills you can give it the ability to see using other models