Speech Recognition Polyfill

A Web Speech API polyfill that swaps webkitSpeechRecognition for local Whisper and Cloud based AssemblyAI transcription.

All local AI Models used are free and don't require any major configuration.

Tested with Duolingo and Google Translate. May work decently with other sites that utilize the API..

Important

The following is accessed for patching sites that use webkitSpeechRecognition. Each serves a specific purpose:

Permissions	Reason
<all_urls>	Inject the content script/polyfill on any page using speech.
storage	Save defaults and per-site overrides (engine, model, language, timeout, debug, cache, etc.).
tabs	Open options page on install and manage icon state with active tabs.

Installation Process

Browser	Installation Steps
	Recommended: Mozilla Add-ons Store - Click Add to Firefox - ✅ Done - ⭐ Rate the addon Alternative (dev build): - Download the latest ZIP from Releases - Go to about:debugging#/runtime/this-firefox - Click Load Temporary Add-on… and pick manifest.json (or the ZIP) - ✅ Done - ⭐ Pin the mic icon to see status colors

Frequently Asked Questions + Models Available

Q	A
How do I make an API Key?	Click the link, create an account, and you’ll get a key right after signing up.
Is the cloud model paid?	AssemblyAI provides a free tier (IIRC, 300–500 hours/month). Beyond that, you have to pay or switch to the local model.
Does audio leave my device?	Local (Default): No audio stays on-device (after the model downloads). Cloud (AssemblyAI): Yes, audio is uploaded for transcription.
Can you explain the icon indicators?	Color reflects recording/processing/error; badges show downloading/cached/done/cancel. A red/error icon often means canceled, missing API key, or unintelligible speech...Not necessarily a bad mic. Pin the icon to monitor state.
How do I improve accuracy?	Speak loud, slow, and clear; pick the correct mic. Use a larger Whisper model (slower) or switch to the cloud engine for better speed & quality.
How is silence handled?	Adaptive Voice Activity Detection plus a configurable silence timeout (global and per-site).

Engine	Model ID	Notes
Local Whisper	Xenova/whisper-tiny.en	English-only, fastest
Local Whisper	Xenova/whisper-tiny	Multilingual, fast
Local Whisper	Xenova/whisper-base.en	English-only, balanced
Local Whisper	Xenova/whisper-base	Multilingual, balanced
Local Whisper	Xenova/whisper-small.en	English-only, higher quality (slower)
Local Whisper	Xenova/whisper-small	Multilingual, higher quality (slower)
Local Whisper	Xenova/distil-whisper-medium.en	English-only, distilled medium (larger/slower)
Cloud	AssemblyAI (API key required)	Remote transcription; model managed by AssemblyAI

Extra Tips

• For learning sites (e.g., Duolingo): set the site language to the one you’re practicing for better speech recognition.
  • For Google Translate, auto-language usually suffices since the site gives us information.
• It does not support continuous speech.

Contributing

• Keep permissions minimal.
• Please organize code.
• Use Debug Mode.

Support Me

Thanks i need it