Gemini Voice Mic

By PRYSKY Studio · Clean, lightweight one-click microphone button for Antigravity IDE, VS Code, Cursor, Windsurf, and other VS Code-based editors.

Free & Subscription-Free (BYOK — Bring Your Own Key): Add your own Google Gemini API key (Free-tier from Google AI Studio or Paid Free+) in 20 seconds — zero Python, zero local CPU load, 100+ spoken languages, and automatic AI chat focus (no need to place the cursor first).

Quick Tips:

• Stackable 1-Minute Takes: Speak up to 01:00 per take. Click 🎤 Voice again anytime to append another paragraph into the same chat box before sending.
• Button Language (10 UI Languages): Automatically matches your IDE language (en, ru, es, pt-br, zh-cn, ja, ko, de, fr, tr), or switch manually via ⚙️ → Extension Settings → Gemini Voice Mic: UI Language.

Key Features

• Zero Cursor Friction: Click 🎤 Voice on the bottom-right Status Bar, speak, and click Stop. The extension immediately focuses the AI Agent chat box and appends your speech.
• Clear 60-Second Take Progress & Live Stopwatch: Inline progress bar (◉ ▰▰▰▰▱▱▱▱▱▱ 00:24/01:00 Stop) with calm coral glow during recording, solid crimson alert in the final 10 seconds, automatic save & insert at 01:00, and one-click Retry if your network hiccups (your recorded voice is never lost).
• Works with Every Chat Model: Transcription converts your voice into clean punctuated text inside the chat input without touching your selected chat model — use it seamlessly with Gemini, Claude, GPT, or any agent model.
• 100+ Spoken Languages & Smart Tech Vocabulary: Speak naturally in English, Russian, Spanish, Portuguese, Hindi, Chinese, Japanese, Korean, German, French, Turkish, or mixed code-switching — programming identifiers (manifest.json, Docker, API, Git, PRYSKY) stay cleanly in Latin script.
• Zero Bloat & 100% Private:
  • On Windows, uses native built-in winmm.dll (no Python or heavy downloads required).
  • Your API key is stored strictly inside your IDE's encrypted SecretStorage.
  • Audio is sent directly over HTTPS to generativelanguage.googleapis.com and cleaned up from %TEMP% after insertion.

30-Second Setup

1. Create a free API key at Google AI Studio (AQ... or AIza...).
2. Click 🎤 Voice in the bottom-right Status Bar of your IDE and paste the key once.
3. Click 🎤 Voice (or press Ctrl+Alt+Space / Cmd+Alt+Space), speak (up to 1 minute per take), and click Stop. Click again to append more takes before sending.

Commands & Settings

• Gemini Voice Mic: Start / Stop Dictation (Ctrl+Alt+Space / Cmd+Alt+Space)
• Gemini Voice Mic: Set or Change Google Gemini API Key
• Gemini Voice Mic: Remove Saved Google Gemini API Key
• Settings (geminiVoiceMic.*):
  • geminiVoiceMic.uiLanguage: auto (default) or choose manually from 10 languages (⚙️ → Extension Settings).
  • geminiVoiceMic.target: autoChat (automatically focus AI chat input) or cursor (insert at current cursor position).

Credits

Developed and maintained as a free community tool by PRYSKY Studio. Licensed under the MIT License.