• Voice recognition?
• At this time of year?
• At this time of day?
• In this part of the country?
• Localized, entirely within your phone?

Yes, and you may see it:

VOSK4Tasker

VOSK4Tasker uses Alpha Cephei's Vosk speech recognition toolkit, and glues it for use with Tasker.

Downloads

The APK is ~50 MBs. Be careful with your data plan!

Building

1. Run ./scripts/english.sh (change english with a different language if exists)
2. Build under Android Studio
3. profit

Manual language model download

If you're not under Linux, or just want to download a different model, here's how to download the models manually.

1. Go to https://alphacephei.com/vosk/models
2. Pick a small model (example: vosk-model-small-tr-0.3)
3. Download zip and extract
4. Rename the folder inside zip to model
5. Place it as app/src/main/assets/model

Usage

After installation, Tasker should recognize this and give you an option to add a new action from the "Plugin" menu. In the action, set "Prompt" to what you want the user to see when asked for speech input, and you're done. The variable %text is your speech to text output.