Amical
A local-first AI dictation app that transcribes speech with Whisper and formats the text for the app you are typing in.GetMIT
Free, open-source alternatives to Wispr Flow
A dictation app that turns speech into cleaned-up text in any app, removing filler words and adding punctuation and formatting.
Wispr Flow is a dictation app: you hold a shortcut, speak, and it types the result into whatever app has focus. Before inserting the text it removes filler words and repetitions, adds punctuation and formats the writing so it reads as typed rather than spoken. Its code is not open source, and your audio is sent to Wispr's servers to be transcribed.
The apps below do the same core job, typing your speech into any app from a shortcut, and several of them can do it without the audio leaving your Mac. VoiceInk, Handy, FluidVoice and Amical transcribe with on-device speech models, while OpenWhispr, Whispering and TypeWhisper let you choose between local models and cloud services. FreeFlow sends audio to Groq or another provider you configure, which is closest to how Wispr Flow works. Several of them can also run the text through an AI model to clean it up before it is pasted.
Apps That Replace Wispr Flow
8 free, open-source alternatives
FluidVoice
A dictation app that turns speech into text in any app, using on-device speech models and optional AI cleanup.GetGPL-3.0
FreeFlow
Hold a key to dictate into any text field, with AI transcription and cleanup through Groq or another provider.GetMIT
Handy
Hold a shortcut, speak, and have offline speech recognition type the words into whatever app you are using.GetMIT
OpenWhispr
Dictate into any app with a hotkey using local Whisper or Parakeet models or cloud services, plus meeting notes and an AI agent.GetMIT
TypeWhisper
Dictate into any app with on-device speech recognition or cloud APIs, then reshape text with workflows.GetGPL-3.0+
VoiceInk
Dictate into any app with on-device speech models, push-to-talk shortcuts and per-app modes.GetGPL-3.0
Whispering
Dictate with a shortcut, transcribe through local or cloud speech models, and paste the polished text at the cursor.GetAGPL-3.0+
