You talk.
It types.

Offline dictation that writes what you meant, even when you change your mind mid-sentence. Free, private, and fast on any PC.

Download for Windows Download for Linux

Windows link opens the Microsoft Store. Prefer a setup file? Get it from GitHub.

You say
Ivy types
Listening Slack

See it work.

Hold a key, talk, let go. Clean text lands in whatever app you were typing in.

Changed your mind mid-sentence?

Most dictation apps type every word and leave the cleanup to you. Ivy's speech model hears you and writes the final version in one pass.

“Let’s meet on Tuesday, sorry, I mean Thursday.”
Let’s meet on Thursday.
“Book a table for four, actually make it six.”
Book a table for six.
“Tell Sam the call is at 3, no wait, 4 PM.”
Tell Sam the call is at 4 PM.
“Um, so, I I think we should, uh, ship it.”
I think we should ship it.

Built for how people actually talk.

The right tone in every app

Casual for chats, Standard for everyday writing, Professional for formal mail. Pick a tone per app and Ivy switches by itself.

Slack Casual

Yeah, I'm gonna push that fix before standup
Thanks!

Notion Standard

Yeah, I'm going to push that fix before standup. Thanks!

Outlook Professional

Yes, I am going to push that fix before standup. Thank you.

Numbers, typed right

₹500 at 3 PM
five hundred rupees at three pm

Money, times and dates the way you'd type them. Big amounts read as 18 lakhs, 2.5 crores or 2 million.

Hold to talk

Alt+Space

Or double-tap for hands-free dictation up to 5 minutes.

Your words, your spelling

Teach it names and jargon. Say a snippet trigger and your saved address or signature appears.

Touch Up

One click after a paste fixes spelling. It never rewrites your sentences.

Polite with your PC

Sleeps while you game or watch full-screen video. Moves to the CPU on battery.

100% offline.

Your voice never leaves your computer. No account, no subscription, no servers listening.

No internet after setupIvy downloads its speech model once. After that it only goes online if you press Check for updates.
History deletes itselfRecordings and transcripts are removed after 24 hours, or up to 7 days if you choose.
Kept out of clipboard historyIvy's pastes skip Win + V history and clipboard cloud sync.
Every line is publicThe full source code is on GitHub, so anyone can check what Ivy does.

One small model. No cloud.

Ivy runs a fine-tune of Qwen3-ASR-1.7B through llama.cpp on your own hardware. It hears and cleans up in a single pass, with no second AI rewriting your words.

1.7B

parameters, fine-tuned on real human speech and dictation paragraphs

0.33s

to write a 15-second dictation on a GPU (any Vulkan card)

3.2s

for the same dictation on the CPU alone

0bytes

of your audio sent anywhere, ever

Rust + Tauri llama.cpp with Vulkan Qwen3-ASR-1.7B fine-tune NVIDIA, AMD, Intel or CPU

Try Ivy today.

Windows 10 and 11, or Linux (.deb and .rpm). Free to use, forever.

Download for Windows Download for Linux

Windows link opens the Microsoft Store. Prefer a setup file? Get it from GitHub.

irm https://raw.githubusercontent.com/raj-7676/ivy-dictation/main/install.ps1 | iex

Or install on Windows from PowerShell with one command.

On Linux, open a terminal and paste the line for your system.

wget https://github.com/raj-7676/ivy-dictation/releases/download/v0.2.9/Ivy_0.2.9_amd64.deb && sudo apt install ./Ivy_0.2.9_amd64.deb

Ubuntu 22.04+, Linux Mint 21+, Debian 12

curl -LO https://github.com/raj-7676/ivy-dictation/releases/download/v0.2.9/Ivy-0.2.9-1.x86_64.rpm && sudo dnf install ./Ivy-0.2.9-1.x86_64.rpm

Fedora 37+

Then open Ivy from your apps menu and hold Ctrl + Shift to talk. Ivy works fully on X11 desktops (Linux Mint, or "Ubuntu on Xorg" at the login screen). On Wayland, the default on Ubuntu and Fedora, Ctrl + Shift is the only hotkey and support is still being tested. Not available yet for Arch, ARM or 32-bit PCs. Full Linux guide.