Open source · Free core
v0.2.0 Latest
  • + Multi-language auto-detection
  • + Recording history with search
  • + Custom text cleanup prompts
  • ~ Improved audio capture quality
View all releases →

Speak.

Dictto turns your voice into clean, ready-to-use text. No data collection. Just talk.

Download for Windows View on GitHub

Windows 10+ · 4MB download · No account needed

Dictto — Recording Demo

Hold Ctrl+Win → speak → release → done

Features

Built for people who ship

System-wide Integration

Works in every text field on your machine. IDE, chat, email, browser — everywhere your cursor blinks.

VS Code
Discord
Slack
Gmail
Notion
+ every text field

4x Faster Input

Speaking is naturally faster than typing. Dictto cleans and formats your words in real-time.

Typing ~40 wpm
Dictto ~160 wpm

Privacy First

Your audio is never stored. Zero telemetry, zero analytics. Everything is verifiable in the source code.

Audio → API → Text → Paste
Nothing stored. Nothing trained.
Verify in source →

Bring your own key

Setup takes 30 seconds

Paste your OpenAI API key. That's it. Your key stays on your machine, encrypted locally.

Interactive demo
Settings → API
sk-proj-....
Configured & Encrypted Locally

Your key never leaves your machine. Zero-knowledge architecture. Verify in source →

Cost breakdown

Your key, your savings

Use your own OpenAI API key. No middleman — you pay OpenAI directly.

Subscription model

$15
/month

Fixed monthly fee.
Data routes through
their servers.

Your own key

~$3
/month

Direct to OpenAI.
You pay only for
what you use.

Questions?

Is it really free?

The core app is open source and free — no word limits, no subscriptions. You bring your own OpenAI API key (~$1-3/month paid directly to OpenAI).

Will Dictto always be free?

The desktop app is open source and free with your own API key — that will never change. We’re building optional cloud features for those who want zero-config convenience. Your current setup stays the same.

Is my data safe?

Your audio is never stored. No telemetry, no analytics, no screenshots. Everything is verifiable in the source code.

Windows shows a warning when installing?

Normal for new apps. Click "More info" → "Run anyway." The code is 100% open source — audit it on GitHub.

What languages does it support?

50+ languages. English, Spanish, French, Japanese — even mix languages in the same sentence.

How does it compare to paid tools?

Subscription tools charge $8-15/month and route your voice through their servers. With Dictto, you use your own API key — open source, direct to OpenAI, and you control the cost.

Stop typing,
start building.

Download for Windows View source on GitHub

Open source. Free core. No account needed.