LOCAL VOICE TYPING FOR WINDOWS AND macOS

Speakintowhat’snext.

Speak your emails, messages and prompts straight into your apps. Yapper turns your voice into text on your PC, with a compact speech model built from Whisper.

Buy Yapper
YAPPER IN ACTION

You talk. It’s written.

Windows and macOS · Local dictation · Designed for CPU use too

LOCAL DICTATION FIRST

Your voice.Your words.Your PC.

The first release starts with speech-to-text on your PC.
Translate when you need to. Explore rewriting with a model that can handle it.

Choose how you speak

Voice activation

Let voice activity control when speech is captured.

Speak freely

Keep the microphone on while you talk, then stop when you’re done.

Push to talk

Hold a key to speak. Release it when you’re done.

Raw

Local speech-to-text. The focus of the first release.

just brainstorming some ideas rn... what do you think?

Translate

Translate your words. Languages and quality depend on the model.

i’m so excited!すごく楽しみです！

Bullet List

AI rewriting · Results depend on the model.

• launch next week • fix onboarding • share with team

Reorganize

AI rewriting · Review wording and meaning.

can u maybe take a look at this when u have time?Could you take a look at this when you have a moment?

Custom

AI instructions · Small models may struggle.

{ "tone": "fun", "format": "tweet", "length": "short" }

These are illustrative outputs, not performance guarantees. Translation and rewriting use a separate language model. Small local models can miss instructions or change the meaning; always review the result.

YAPPER IN ACTION

You talk. It’s written.

BEYOND DICTATION · MODEL-DEPENDENT

Createyour ownmode.

Give your words a different shape with instructions for a language model. These modes already work, but small local models can struggle to rewrite a whole thought faithfully. The examples show the goal, not a guaranteed result.

  • The model makes the difference.

    Simple instructions suit smaller models. More capable models can follow complex instructions better, but still need a check.

  • Your API key · In preparation

    We’re preparing optional connections to external AI providers. Your text would be sent to the provider you choose, with its own privacy terms and costs.

  • More local models · In preparation

    We’re working on loading additional compatible models, including from Hugging Face. Larger models need more memory and processing power. Supported formats will be confirmed before release.

yapperSettings

Text modes

Mode nameThe team update

Always includedTransform my words. Don’t answer them. Return only the result, without commentary.

Instructions for the model

Rewrite as a concise team update. Keep names and deadlines.

Protect numbers and codes
Make this mode in Yapper
Team updateILLUSTRATIVE EXAMPLE
YOU SAY

“Okay so the login fix is done, tomorrow I’ll test the checkout, and I’m still waiting for Maya’s API key.”

YAPPER TYPES

Done: Fixed login.

Next: Test checkout tomorrow.

Blocked: Waiting for Maya’s API key.

Illustrative results. Small local models may miss instructions or change details. Additional models and API connections are in preparation; review every rewrite.

Your apps. Now with a voice.

EmailsMessagesAI promptsDocuments

FROM THOUGHT TO TEXT

One shortcut. You’re writing.

  1. 01

    Pick your place

    Click a text field in the app you’re using. A message, an email, a prompt. That’s where your words will go.

  2. 02

    Press. Speak.F1

    Open Yapper with Win + F1. The mic is ready. Pick a mode before dictating; change the shortcut in settings.

  3. 03

    Pause. It’s there.

    Raw writes progressively. AI modes insert a finished block at a natural pause. Keep talking when you’re ready.

LOCAL BY DESIGN

Your voice. On your PC.

Local speech-to-text is the heart of Yapper. We’re preparing our own quantized model based on Whisper: a compact version designed to make everyday dictation practical, including on CPUs.

YOUR COMPUTER
Your voiceYapperYour text
Local speech-to-text.

Designed for CPUs too

A dedicated GPU shouldn’t be a prerequisite for dictation. Speed still depends on your processor, available memory and settings. We’ll publish tested requirements with the release.

Go offline

After downloading the app and speech model, local dictation works without internet. Local language models can also work offline; optional external AI providers require a connection.

Keep it small

A compact toolbar, a configurable shortcut, and a quiet place in the system tray. Models are released after inactivity.

Local dictation keeps voice processing on your computer. Planned connections to external AI providers are optional: text sent to a chosen provider leaves your PC. The website and account use online services.

GOOD QUESTIONS

Glad you asked.

A little more about the little bird.

Does Yapper work on my computer?

Yapper supports 64-bit Windows 10 and 11, and macOS 13.3 or later on Apple Silicon and Intel Macs. Local dictation is designed for CPU use too. Speed varies with your processor, memory, model and settings.

Can I use it with my existing apps?

Yapper inserts text into the field you’ve focused, so you can dictate into messages, documents and prompts. Compatibility can vary with the app and its permissions. Change focus deliberately before you start recording.

Is everything processed locally?

Dictation runs on your PC with our speech model based on Whisper. Rewriting also stays local when you use a local language model. Planned optional API connections send text to your selected external provider and may incur its charges. Initial downloads and the website need internet.

Does it translate while I speak?

Translation and reorganization are committed in blocks at natural pauses. You can keep the microphone on between blocks. Yapper doesn’t continuously rewrite text you’ve already typed.

Can I create my own modes?

Custom modes let a language model work from your instructions. They’re more demanding than dictation: small local models may omit details or fail to follow the instructions. We’re preparing support for personal API keys and additional compatible local models, including downloads from Hugging Face. Providers, formats and availability will be confirmed before release.

Can I rely on every AI rewrite?

No model gets every instruction right. Small models can produce a fluent result that still changes or omits something important. Review names, numbers and meaning, and use direct dictation when you don’t need the extra processing.

How much does it cost?

€20 once: a lifetime licence for one person on two computers, with 12 months of updates included. After that, another 12 months of updates cost €12.99, once or renewed annually and cancellable anytime. Without renewing, your licence stays valid. Optional external AI providers may charge separately.

LESS KEYBOARD. MORE YOU.

Your next message?
Just say it.

Buy Yapper

Create your account to use Yapper on Windows or macOS.

LOCAL DICTATION FOR WINDOWS AND MACOS

Yours. For good.

Buy Yapper once.
Updates after the first year are optional.

Available for 64-bit Windows 10 and 11, and macOS 13.3 or later on Apple Silicon and Intel. No recurring fee for local dictation. Requirements and details

BUY YAPPER

€20 one time

Lifetime licence · 12 months of updates included

  • Local dictation on Windows and macOS
  • One person, two computers

OPTIONAL UPDATES

Already own Yapper?

Add another 12 months of updates.
Your licence stays valid even without renewing.

ONE-TIME RENEWAL

€12.99 one time

12 months of updates,
without automatic renewal.

AUTOMATIC RENEWAL

€12.99 / year

12 months of updates,
renewed annually.
Cancel anytime.

11 people joined before launch.