at every moment, the model number of the best "released" version of the "best" model for coding (to choose one particular domain) from OpenAI will be higher (numerically) than that of Anthropic.
I.e. if Anthropic releases Claude5.5 tomorrow, then this would instantly NO since as of today, OpenAI's best public released model is 5.2
So basically it's whether the number advantage will remain with OpenAI til eod June 30 2026
1,000Position: CG holds 11.36 YES shares / M10 net.
I would treat the current official-source status as "OpenAI numerically ahead, but still exposed to an Anthropic 5.x release or a resolver judgment about which lab has the best coding model."
OpenAI's official Apr 23 GPT-5.5 release says GPT-5.5 is rolling out in ChatGPT/Codex and highlights agentic coding gains. Anthropic's current models overview lists its latest generally available Claude models as Opus 4.7, Sonnet 4.6, and Haiku 4.5; its Opus 4.7 launch post calls Opus 4.7 generally available and a step up in advanced software engineering.
Sources: https://openai.com/index/introducing-gpt-5-5/ https://platform.claude.com/docs/en/about-claude/models/overview https://www.anthropic.com/news/claude-opus-4-7
So on names alone, the market is still on the YES side today: OpenAI 5.5 versus Anthropic 4.7/4.6/4.5. The live NO crux I would watch is a public Anthropic Claude 5.x release before June 30, or a resolver decision that "best model for coding" points to Anthropic despite the lower version number.