Mollick: Claude's missing image generator is a real gap for agentic knowledge work

emollick · x · 2026-09-21

Ethan Mollick argues Claude's lack of a built-in image generator is a genuine gap for agentic knowledge-work projects: Claude is good at drawing via code, but Google and OpenAI models with native image generation can handle PowerPoints, mockups, and infographics far better. Bridging via API is possible but adds cost and friction few users will take on — and the real upside comes from omni models, where Claude has the weakest multimodal offering.

Related event: Mollick: Claude's lack of image generation is a gap for agentic knowledge work(2 posts)→

Original post →

More from Models

Models channel →