Anthropic's March 2026 update to Claude Code is one of the biggest yet, introducing voice mode, recurring task automation, and a massive context window expansion.
Voice mode activates with the /voice command and works through a push-to-talk mechanism — hold the spacebar to speak, release to send. Transcription has been optimised for technical terms and repository names, which is crucial for developers who talk about code. It is a natural evolution for CLI-based AI tools that have been text-only until now.
The /loop command transforms Claude Code into a lightweight monitoring system. Define an interval and a prompt, and Claude will execute it on a recurring schedule — essentially a session-level cron job for code health checks, build monitoring, or deployment tracking.
Perhaps the headline feature is the 1M token context window, now available by default for Opus 4.6 on Max, Team, and Enterprise plans. Previously this required extra usage allocation. Combined with the default output token boost to 64K (with a 128K upper bound), Claude Code can now reason over entire codebases in a single session.
