v1.0 Now Available

Hold. Speak.
Release.

Type with your voice anywhere on Windows.Local AI. Zero latency. 100% private.

Get notified about updates and new features

> Injecting text at cursor...

Why VoiceCursor?

Built for power users who demand speed and privacy.

Universal Injection

Types at your cursor position in ANY application. Browsers, IDEs, games, chat apps.

100% Local & Private

Runs entirely on your PC. No audio ever leaves your machine.

GPU Accelerated

Whisper large-v3-turbo on NVIDIA GPU. Sub-300ms processing.

Korean + English

Auto-detects language. Handles mixed Korean/English speech perfectly.

Simple Pricing

No subscriptions. No recurring fees. Pay once, own it forever.

Free

$0/forever

Try VoiceCursor with basic features

  • 10 minutes per day
  • Base model
  • Basic features
  • Community support
Best Value

Standard License

$23.99/one-time

One-time purchase. Own it forever.

  • Lifetime access to all features
  • 1 year of free updates included
  • Unlimited recording time
  • Large-v3-turbo model (most accurate)
  • Real-time translation
  • Priority support

System Requirements

Make sure your system meets these specs for the best experience.

Operating System

Windows 10/11

64-bit, latest updates recommended

Graphics Card

NVIDIA RTX 2060+

6GB+ VRAM required for optimal performance

Processor

Intel i5 / Ryzen 5

AVX2 support required

Memory

8GB / 16GB RAM

16GB recommended for multitasking

Storage

5GB Free

SSD recommended for model loading

Network

Required

For setup & license check only

No GPU? CPU mode is available with the base model (slower processing).

Frequently Asked Questions

Everything you need to know about VoiceCursor.

How does VoiceCursor work?

It's simple: Hold the customizable hotkey, speak naturally, and release. The text is instantly typed wherever your cursor is—whether it's Discord, VS Code, or a game chat.

Is my audio sent to the cloud?

No. VoiceCursor runs 100% locally on your machine using OpenAI's Whisper model. Your audio never leaves your PC, ensuring complete privacy.

What languages are supported?

Currently, we support English and Korean. The model automatically detects which language you are speaking, even if you switch mid-sentence.

Can I use it without a GPU?

Yes, but we highly recommend an NVIDIA GPU (RTX 2060 or better) for the instant 'hold-release' experience. CPU mode works but will have a noticeable delay.

How do I get my license key?

Immediately after purchase, your license key will be emailed to you. You can also view it anytime in your dashboard on this website.

Can I use it on multiple computers?

Yes! Your license allows you to activate VoiceCursor on up to 2 devices simultaneously.