Dictaro · Chrome extension

A few steps to your first take.

Use Chrome 114 or newer on a computer with a microphone. Keep an internet connection for the first model download, and at least 500 MB free if you select the Small model.

1. 1. Download and extract

   Download Dictaro 0.1.0 for Chrome. Extract the ZIP to a permanent folder. It contains manifest.json, the popup, speech engine and third-party notices.

   Download the SHA-256 checksum. On Linux use sha256sum -c dictaro-chrome-0.1.0.zip.sha256; on macOS use shasum -a 256 dictaro-chrome-0.1.0.zip; on Windows use Get-FileHash dictaro-chrome-0.1.0.zip -Algorithm SHA256 and compare the digest.

2. 2. Load the extension

   Open chrome://extensions in Chrome, enable Developer mode, choose Load unpacked, and select the extracted folder containing manifest.json. Pin Dictaro from Chrome’s Extensions menu.

3. 3. Choose your model and language

   Open Dictaro’s popup. Search100 supported languages by English name, native name or code, or choose Auto-detect. Tiny is the fastest option (about 78 MB); Base is about 148 MB and Small about 488 MB. English uses an English model; other languages and Auto-detect use the multilingual model. The first start downloads model data from Hugging Face and caches it in your extension profile.

4. 4. Focus a field and speak

   Refresh the page after loading the extension. Click a text input, textarea or editable text in the main page. Start with the popup or Ctrl+Shift+1 (Command+Shift+1 on macOS). Allow microphone access when Chrome asks. Start becomes Stop while the model or microphone is preparing, so you can cancel. Chrome reserves some shortcuts; change yours at chrome://extensions/shortcuts if needed.

5. 5. Stop, review and copy

   Stop finishes the take and applies your chosen cleanup rules. Dictaro keeps the transcript in its popup. If you switch fields, the page changes or insertion is unavailable, copy the transcript rather than losing it. Chrome system pages and nested iframe editors are not insertion targets.