This multimodal API from @Aleph__Alpha provides text prompt completions using images as context, like Flamingo. In each screenshot, the bold text is the prompt, and the non-bold text is AI-generated:

→ View original post on X — @goodside