Mojo UpDocs

MojoFlow

Push-to-talk dictation for Windows and Linux. Hold a key, speak, let go, and your words are typed where your cursor is. Everything runs on your computer.

MojoFlow turns your voice into text in any app. Hold your push-to-talk key (Right Ctrl unless you change it), say what you want to write, and let go: the words appear where your cursor is, already punctuated and tidied up.

Nothing you say leaves your computer. A speech model on your graphics chip turns your voice into text, and a small language model, also on your computer, removes the "um"s, adds punctuation and applies "no wait, I meant…". MojoFlow works offline once its models are downloaded.

Install MojoFlowDownload the Windows installer, or install on Linux with one command.Your first dictationThe two-minute setup: speech model, key, microphone and a practice box.DictateHolding or tapping the key, your microphone, language and spelling, and how text is typed.Clean-upOff, Light or Polished, and a different style for each app.DictionaryTeach MojoFlow names and words it gets wrong.With AI WorkbenchDictate straight into the workbench chat, and share its model.

What you need

A Windows 10 or 11 PC, or a Linux desktop, with about 4 GB free for the models. A graphics chip makes it fast, but MojoFlow also runs on the processor alone. See System requirements.

MojoFlow is a Mojo Up product. It works on its own, and works better beside AI Workbench if you use both.

Install MojoFlow

Download the signed Windows installer, or install on Linux with one command. No account is needed.

On this page

What you need