The free, open-source alternative to Wisprflow

Speak your mind.
Let AI do the typing.

A free, open-source speech-to-text tool that brings lightning-fast, cloud-powered dictation to every text field. Under 1 second latency.

Available for macOS. Windows support is currently in development.
Open Source
Privacy Conscious
BYOK (BRING YOUR OWN KEY)
Works Across All Apps

Why VoiceFlow?

Your Intelligence

By bringing your own keys (BYOK), VoiceFlow stays completely free. We've curated independent top-tier providers that offer generous free tiers or cost pennies per hour.

Simply plug in two keys: one for lightning-fast transcription, and one for intelligent AI cleanup.

Step 1

Transcription Model
Groq
Deepgram
AssemblyAI

Step 2

AI Cleanup
Gemini
GPT
Ollama
+ Local, Open-Source & Enterprise Models

History & Search

Never lose a thought. Access your previous transcriptions through a secure, local history log.

Technical Precision

A custom dictionary for your specific vocabulary, dev terms, and internal jargon.

Anywhere, Any App

Right from your PC, system-level injection sends polished text wherever you work. Whether you're in Slack, VS Code, or a terminal, VoiceFlow types wherever your cursor lands.

Always Listening

Your Always-On
AI Assistant

Meet the companion that listens perfectly, cleans up your thoughts, and saves you hours of typing every week.

Flawless Grammar
< 1s Latency
?
Clean text pasted
Transcription Failed
Please check your connection

Magic in 5 Steps

The fastest open source path from voice to text ever built.

Hotkey

Trigger recording with your custom combination.

Record

Captures crystal clear audio at 16kHz mono.

Transcribe

Cloud-speed processing via chosen provider.

Clean

LLM fixes errors and removes filler words.

Inject

Text appears instantly at your cursor.

Ready to start?

Get up and running with VoiceFlow in less than 2 minutes. No subscription required, just bring your keys.

1

Install & Open

Download either from GitHub or the download buttons on this website.

2

Grant Permissions

Microphone for audio, Accessibility for pasting, and Input Monitoring for hotkeys.

3

Add 2 API Keys

You need exactly two keys: one for your transcription provider and one for your cleanup provider, chosen from the options we offer.

4

Set Your Hotkey

Choose between "Hold-to-record" or "Toggle" modes for maximum flow workflow.

Built like a product assistant

Free for life.Private by design.

Clear economics and clear boundaries—so you always know what leaves your PC and what stays yours.

BYOK onlyNo telemetryLocal history

Transparent cost

VoiceFlow costs you nothing upfront. You only pay providers for what you use with your own keys—no subscription, no markup, no middleman.

No recurring product fees

Your data, your keys

Audio and text go straight from your PC to the APIs you choose. We don't see it, store it, or train on it.

You choose every provider

Frequently Asked Questions

Is it free?
Yes, the software is free and open-source. However, you handle your own API costs directly with providers like Groq or Deepgram (which often have generous free tiers).
Does it work offline?
Currently, VoiceFlow requires an internet connection to reach cloud APIs for transcription and LLM cleaning. Local Whisper support is on our roadmap.
What about privacy?
VoiceFlow relies on your chosen provider's privacy policies. Since you use your own API keys, you have full visibility and control over your data agreements.
Is there a Windows version?
We are actively working on a Windows version using .NET and Windows Accessibility APIs. Stay tuned to our GitHub for updates!
What permissions are needed?
You'll need to grant Microphone (to hear you), Accessibility (to paste text into other apps), and Input Monitoring (to detect your hotkey globally).

Source & releases

Get VoiceFlow

Completely open-source. Clone or fork on GitHub today—installers for macOS and Windows land here as soon as they're ready.

View on GitHub

Installers