Someone called it AI's game of thrones And it's very interesting indeed, you're seeing model providers trying to shapeshift into frontends and vice versa It started with model providers like OpenAI creating big LLMs And then OpenAI launched ChatGPT, so they became a model provider + a chat frontend Some ppl left OpenAI and founded Anthropic which made Claude, a model + chat frontend, although I don't know many people who actually use their chat frontend, people rate them for the quality of their coding models mostly Then Cursor launched and you had the first popular coding frontend which could use any model provider in the back (like OpenAI, Anthropic, Google etc.) But then Anthropic made their own coding frontend recently called Claude Code, that became very successful Meanwhile OpenAI tried to build their own coding front end first with Codex, but it wasn't so good, so then they tried buy one with Windsurf, but that bounced off -- So now you have this funny strategic positioning of AI companies: OpenAI: great models, most popular chat frontend, no real coding frontend Anthropic: great models, non-popular chat frontend, a very popular coding frontend Cursor: no models, n/a, a very popular coding frontend Google, great models, non-popular chat frontend, no coding frontend Very fun to see this stuff play out like some chess game