Quick start
1. Download and install
- Open the VoicePaste download page and choose your operating system.
- On macOS, choose Apple Silicon or Intel to match your Mac. On Windows, choose the x64 installer.
- Finish installation, open VoicePaste, and go to Settings.
Windows users: Installation may trigger a system security notice. See Common questions at the end of this page for help.
2. Grant system permissions
- Open Settings → System Permissions and use the microphone check button. Allow access when prompted.
- On macOS, also grant Accessibility permission so VoicePaste can paste text into other apps.
- Return to VoicePaste and check that the required permissions show as granted.
See System permissions for the full steps.
3. Configure speech recognition
A speech recognition model (ASR) converts audio into text. Choose either an online or a local model.
- Online: Get an API key using the Volcengine (Doubao) or ElevenLabs guide. In Settings → Audio Models → Online Models, enter your credentials, save, and enable the model.
- Local: In Settings → Audio Models → Local Models, download a model and enable it when ready. No API key is needed.
See Recognition models for setup and the compatibility list for supported features.
4. Configure a language model (optional)
For translation or polishing, configure a provider under Settings → Large Language Model and add a text template. Skip this step if you only need transcription.
Start with the DeepSeek API guide, then follow Language models and templates.
5. Set shortcuts and try your first input
- Open Settings → Shortcuts, check the bindings, and choose hands-free or push-to-talk mode.
- Open a text editor, focus an input field, and start recording with your shortcut.
- Say a sentence, finish recording using your chosen mode, and check that the text appears.
See Shortcuts and input.
Common questions
What if Windows says “Windows protected your PC”?
Confirm that the installer came from this site's download page, then choose “More info → Run anyway.” For other blocks or firewall prompts, see System permissions: common questions.
What if no text appears after recording?
Check system permissions, the enabled recognition model, and whether the destination input has focus. See System permissions and Recognition models for troubleshooting.