Model Releases
Claude Code is tuned for Claude. Codex is tuned for OpenAI models. Until now, ππππππππππ had fixed harness defaults, which meant iβ¦
Claude Code is tuned for Claude. Codex is tuned for OpenAI models. Until now, ππππππππππ had fixed harness defaults, which meant it couldn't take advantage of the provider-specific optimizations that
Claude Code is tuned for Claude. Codex is tuned for OpenAI models. Until now, ππππππππππ had fixed harness defaults, which meant it couldn't take advantage of the provider-specific optimizations that make those systems work so well Profiles change that! You can tune prompts, tools, middleware, skills, and more for whichever model you're running. The gains are pretty huge: 10β20 point improvements on harder tau2-bench tasks π We also ship built-in profiles for OpenAI, Anthropic, and Google, so you don't need to start from scratch
Source: Harrison Chase (X) | 2026-04-30