Superwhisper cloud dictation is the fresh-install default in merged code
Pull request 14625 merges Superwhisper cloud dictation as the fresh-install default, with no install prompt and no offline model download. Pull request 14552 merges selectable Voxtype and Superwhisper dictation. Neither merge is in version 4.0.4.
The merged change leaves an existing dictation choice as it is. Pull request 14552 says Setup > Defaults > Dictation installs the selected backend when it is missing. It says holding Right Alt or F9 dictates, and Super + Ctrl + X toggles.
The fresh-install package list in pull request 14625 includes superwhisper-bin, and the saved default file contains superwhisper. Pull request 14552 writes voxtype into that file only when it is missing and voxtype-bin is already installed. The merged shell drops its own dictation indicator, and the provider draws the recording overlay.
David Heinemeier Hansson opened and merged both changes. He wrote that no audio was recorded and that a fresh virtual-machine install was not run.
Pull request 14625 says the package database lists superwhisper-bin 0.1.0beta15-1 for x86_64. The newest Omarchy release is version 4.0.4, so these merges are not part of that release.