Simply put, I want to know, today, which should I be using?
The "code-supernova-1-million" stealth model was more capable, but it is unfortunately no longer available through any of providers. Hopefully it gets publicly released soon. My guess is that it is an updated code model from XAI/Grok.
Having to switch models every few weeks isn't the most fun, but I can't turn down the productivity that comes with it. I wish I could just go back to Neovim, but I can't justify the lower productivity.
ChatGPT – journaling, talking, planning. Codex – framework and middleware-layer coding. Claude Code – logic and application-level coding. Anthropic models via OpenRouter + Cline – when the task is error-prone, tedious, or needs high fidelity; lower error rate in my experience, though pricier. Cursor Agents – multi-file integration, boilerplate, and forking tasks.
Each fills a different slot in the workflow, so “best” depends on what kind of coding you’re doing.