Anthropic recently throttled Claude usage due to demand. The linked article, among others, speculate that it is due to political reasons, because CEO Dario Amodei refused the demands of the Pentagon. I actually believe that their surge in popularity comes from a more durable source, and that is their brand.

I use Claude Code at work and Gemini CLI in my personal projects. It is hard to make an apples-to-apples comparison, since the codebases are different in size and complexity, but in my personal testing, I have not found a meaningful difference in output quality. I imagine that trying Codex would yield similar results.

The main difference I do feel, however, is the branding difference. Some of it may be due to the recent political event, but Claude has always had a more human-centric feel to it. Obviously, Claude’s name can actually refer to a person, as opposed to Gemini or GPT-4o. Even versions of Anthropic’s models are given more legible names, e.g. “Sonnet” or “Opus”.

Perhaps more subtly, the brand colors are warm and gentle, vs. tech future focused.

Claude is also the only one with a serif font, which is not only more nostalgic, but also establishes more continuity with humanity’s written history.

It is too early to tell if this brand distinction will remain a competitive advantage. Only time will tell.