Claude Code 2.1.284 makes Sonnet 5.5 default and shows cost in dollars
Anthropic released Claude Code 2.1.284 on September 29, switching the default model to Sonnet 5.5 and surfacing dollar-denominated run cost in the terminal.
AI/TLDR reported on September 29 that Anthropic released Claude Code 2.1.284, switching the terminal agent’s default model to Sonnet 5.5 on the same day the new model shipped, and starting to show run cost in dollars in the Claude Code terminal. This is the first time Claude Code has surfaced cost information directly in the user interface, letting users see per-task token spend as the coding session progresses.
The release also bundles a set of performance and stability updates, continuing the 2.1.280-2.1.283 line shipped September 25, focused on fixing memory pressure during long sessions and hardening the agent tool-call chain. Anthropic had already telegraphed that Sonnet 5.5 is the preferred default for Claude Code, citing parity pricing with Sonnet 5 but more than 30 percent faster task completion.
Claude Code 2.1.284 also improves compatibility with third-party CLI orchestrators such as OpenRig, which lets users run Claude Code alongside Codex in a single YAML-defined team. This means OpenRig-style agent orchestration layers can treat Claude Code and Codex as one managed team.