Skip to content

[FEATURE] Let users replace words Apple Speech keeps misspelling - #124

Open
JulianPscheid wants to merge 1 commit into
tornikegomareli:mainfrom
JulianPscheid:feature/spelling-replacements
Open

JulianPscheid wants to merge 1 commit into
tornikegomareli:mainfrom
JulianPscheid:feature/spelling-replacements

Conversation

@JulianPscheid

@JulianPscheid JulianPscheid commented Sep 4, 2026 •

Copy link
Copy Markdown
Contributor

Apple Speech cannot be taught a name it does not know, so a word like Hedy lands as Hetty in every session. This adds a Spelling replacements list in Settings: from-to pairs applied after recognition and before Prompt Shaping or translation, and on Drop Transcription the same way. Matching is whole-word and case-insensitive, the left side can be a phrase for a name the recognizer splits up, possessives keep their 's, and an empty list is the previous session. I used Grok and Codex while building and reviewing this.

Tested on a 16-inch MacBook Pro (M4 Max, macOS 26.6.2) with the built-in notched display, adding and editing pairs in Settings and dictating real sessions to watch the replacements land.

Image

Closes #119

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Direct Dictation has no way to correct a name it always misspells

1 participant