Official documentation
DoNotType: README (GitHub)
bojieli, GitHub (bojieli/DoNotType), 2026
Read the originalRead on October 22, 2026
- What it is
- The README of DoNotType, an MIT-licensed voice dictation app for macOS, Windows, Android and iOS, in the maker's own GitHub repository. It describes its modes, where audio is sent, its personal dictionary and its principles on accounts and servers.
- How much weight it carries
- The maker's own repository page, read on 2026-10-22. It is the maker's own description. The README names the provider the user chooses rather than a DoNotType server, and does not say whether recordings are used for training, so no row states that.
Facts taken from this source
- DoNotType's README says it has a local dictionary with one-column CSV import.
“A local dictionary with one-column CSV import; correction learning is opt-in, labelled, and removable” (Highlights section, table row Personal dictionary)
- DoNotType's README says there is no DoNotType server in the middle, so audio does not go to the maker's servers.
“No DoNotType server in the middle. There is no DoNotType account, sign-in, subscription, telemetry, or analytics.” (Principles section, item 3)
- DoNotType's README says its Rewrite mode can make the result formal, concise, or casual.
“Rewrite can make the result formal, concise, or casual.” (Dictation that stays yours section, first paragraph)
- DoNotType's README says audio and screen context go directly to the AI provider the user chose.
“Audio and context go directly to the provider you chose.” (How it works section, table, column 3 Transcribe)
- DoNotType's README says audio and context go directly to the provider the user chose.
“Audio and context go directly to the provider you chose.” (How it works section, table, column 3 Transcribe)
- DoNotType's README says the text is inserted at the cursor in the chosen style.
“The text is inserted at the cursor in your chosen style.” (How it works section, table, column 4 Insert)