Hold Right Shift
Speak naturally, then release. VoiceFlow transcribes with the bundled offline speech model and pastes into Chrome, Notion, docs, email, chat tools, and more.
VoiceFlow 3.1.0 for Windows
VoiceFlow V3.1
A private Windows desktop assistant for voice dictation, AI writing, screenshot questions, and real work that stays in flow.
Windows 10/11 | Direct EXE download | Groq key required for AI features
Offline dictation
Local-first history
Screen questions on demand
Right ShiftHold to dictate, release to paste.
Ctrl + SpaceHold to speak an AI command.
Screen demo: voice dictation, mouse push-to-talk, and AI commands in one flow.
AI video slot
Explore VoiceFlow below
Write without breaking focus.Dictate into the app you already have open.
Ask about what is on screen.Crop a screenshot and get direct, useful help.
Keep control of your data.History and screenshots stay on your Windows PC.
Voice-first workflow
Use AI without leaving the app you are in.
Hold a shortcut, speak naturally, then release. VoiceFlow handles dictation and AI commands without interrupting your workflow.
Hold Ctrl + Space
Speak a request, then release. VoiceFlow can generate text, rewrite clipboard content, answer questions, or open sites like YouTube, Google, Notion, and ChatGPT.
What changed in V3
From voice dictation to desktop assistant.
V2 focused on dictation, AI commands, and clipboard rewrites. V3 adds a fuller assistant layer around the orb.
Floating voice orb
An always-available desktop orb with recording, thinking, typing, and done states plus a right-click menu for app controls.
AI text cleanup
Clean dictated text with grammar, punctuation, capitalization, and filler-word fixes while preserving your meaning and tone.
VoiceFlow AI chat
Use a fixed right-side AI chat panel for normal assistant conversations without opening ChatGPT separately.
Screenshot + ask screen
Capture your screen, crop the useful part, attach it to chat, and ask what it means or what to click next.
Diagnostics and reconnect
Check API key status, AI connection, speech model, microphone, hotkeys, and the last error from one panel.
Local history
Keep recent generated results available from the orb menu, copy previous outputs, or clear history when you want.
Assistant tools
Talk to your computer, your text, or your screen.
The V3 assistant layer is built for everyday desktop moments: writing a message, understanding an error, opening a web app, or turning an idea into content.
Ask the screen
Use screenshot crop and vision AI for questions like "What is on my screen?", "Explain this error", or "What should I click next?"
Generate and rewrite
Speak or type prompts to write messages, create drafts, summarize copied content, translate text, or improve tone.
Open common sites
Use voice commands to open YouTube, Google, Notion, ChatGPT, and other everyday web apps faster.
Setup guide
Install and start speaking.
The current shareable build is a direct Windows EXE. Setup is straightforward, with one expected Windows warning while the app remains unsigned.
Download VoiceFlow V3.1
Download VoiceFlow.exe, then run the Windows app.
Allow the beta app
Windows may show Unknown publisher. Choose More info, then Run anyway. Code signing is planned for a wider public release.
Choose your setup
Dictation uses the bundled offline speech model. For AI chat, commands, cleanup, and screenshot questions, enter your Groq API key.
Quick controls
Keep these shortcuts close.
- Right Shift Hold to dictate, release to paste.
- Ctrl + Space Hold to speak an AI command, release to run it.
- Backspace Cancel an active recording.
- Escape Press twice quickly to quit VoiceFlow.
- Right-click the orb for settings, AI chat, diagnostics, history, and app controls.
Beta notes
What to expect in 3.1.0.
- Windows 10 or Windows 11 is required.
- Dictation is English-only for now.
- Hotkeys may not work inside apps running as Administrator.
- App history and screenshots are stored locally on your Windows PC.
- AI features send only your selected prompt, text, or cropped screenshot when you choose to use them.
Try these AI commands
Speak, type, or attach a screenshot.
Screen
"Explain this error on my screen."
Crop a screenshot, attach it to chat, and ask VoiceFlow what is happening.
Create
"Write a clear reply to this message."
Copy a message or dictate context, then let VoiceFlow draft something you can paste back.
Clean up
"Make this message cleaner but keep my tone."
Use AI cleanup or the built-in chat panel to polish copied or dictated text.
Available for work
Looking for the next opportunity.
I build practical software products like VoiceFlow and I am currently open to work.
Hire me
Open to jobs and freelance projects.
I work on Python tools, Windows apps, AI integrations, and automation. If you are hiring, let us talk.
Funding and traction
Help turn VoiceFlow into a real product.
I am tracking early traction and looking for support, mentorship, and opportunities to build VoiceFlow into a polished real-world app.
37
Total GitHub downloads as of now
v3.1.0
Current shareable Windows build
Live
Beta feedback form ready for testers
What funding helps build next
- A more structured VoiceFlow version with cleaner onboarding, settings, and update flow.
- Upgraded AI models for better writing, faster answers, and stronger screen understanding.
- Deeper AI conversation features so VoiceFlow feels more like a real desktop assistant.
- More next features around voice commands, productivity workflows, and everyday desktop help.
Interested in supporting or hiring?
If you are an investor, mentor, recruiter, or founder who likes practical AI tools, I would love to talk.
Optional donation
If VoiceFlow helped you, you can support the next products.
No pressure at all. If you liked this project and want to help me build more useful AI tools, a small UPI donation helps with APIs, testing, hosting, and future releases.
VoiceFlow 3.1.0
Ready to test the Windows beta?
Try dictation, AI chat, screenshot questions, website opening, cleanup, diagnostics, and history, then share what felt smooth and what got in your way.
Beta feedback
Tell me what worked and what got in your way.
Your notes will help shape the next VoiceFlow build. Short feedback is useful too.