Ampersand
0.4.4

What changed

• Speech-to-text models mishear names, technical terms, and word boundaries, especially outside English. A new setting in Settings → Meetings (“Correct the transcript first”) runs your chosen AI provider over the raw transcript in chunks before writing the note, so the summary is built from what people actually meant. Each chunk carries context from the previous one to keep names consistent across the recording. The note’s provenance line clearly discloses when wording was corrected by AI rather than quoted verbatim.
• Toggles in published notes and HTML exports now render as native HTML <details> elements. Meeting transcripts and other collapsible sections stay neatly folded until a reader clicks to expand them, with zero JavaScript required. When exporting to PDF or printing, toggles still flatten completely so no text is lost on paper.
• On iPad, Ampersand introduces a Mac-style tabbed note workspace with multi-tab switching, document navigation, and refined touch table controls.

Ampersand updates itself in place, so you are probably already on a newer build than this one. If you are installing fresh, the download page always has the current release for macOS, Windows and Linux.