Voice-Controlled Coding Assistant

This is my take on reimagining how I write code: ditching the keyboard (if you let it) and letting your voice do the heavy lifting. Inspo from teacher who recommended something for physically impaired coders.

For v1.0, I kept it relatively simple:

• A landing page that shows you requirements + instructions
• A main coding interface with an editor that feels familiar if you’ve ever touched VS Code
• LLM status showing either enabled or disabled (make sure to put it in API key on the top-right)
• File rename options
• Toggle text wrap
• Artifical Intelligence, what I call, smartness. This uses an API from OpenAI's OR Groq's model and you can either set it to "strict" or "smart", the difference being in how accurately it will execute your actual prompt. Strict will generate code based exactly on what you said. For instance, if you said "write a print statement that will print the final score", the output would look something like this:

print(finalScore)

If you set the smartness to "Smart", then the output would look like this given the same case:

print(f"The final score is {finalScore}")

The idea is to get the vibe right first with smooth animations, pulsing indicators when it’s listening, a layout that feels like an actual dev tool and not a clunky prototype. From here, it’s not a huge leap to plug in real AI models, adaptive syntax learning, and proper debugging commands.

Getting Started

AI Setup

VoiceCode uses your own API key for AI features. This keeps your key private and lets you control costs.

Quick setup: Click the Key icon (⚿) in the right side of the top bar and follow the instructions. I recommend Groq (free, fast). See USER_GUIDE.md for detailed setup.

Prerequisites

• Node.js (v18 or higher)
• npm (comes with Node.js)

Installation and Setup

1. Clone this repo:

   git clone https://github.com/aryanvx/voice-coder-studio.git
   cd voice-coder-studio

2. Install dependencies:

   npm install

3. Start the dev server:

   npm run dev

4. Check port allocated in the terminal and open it in your browser

Optional: Enable LLM-based code generation

If you want the editor to generate full function implementations using an LLM (OpenAI), create a .env file in the project root with:

VITE_OPENAI_API_KEY=sk-xxxxxxxxxxxxxxxx

or

GROQ_API_KEY=gsk_xxxxxxxxxxxxxxxx

Then restart the dev server. When enabled, the editor will show an "LLM: Enabled" indicator and preview any generated suggestion before you accept or reject it.

This is just the start.

The plan is to turn this into a full AI-poIred development environment that can understand your coding style, explain your bugs, and maybe even argue with you about why your variable names are terrible (seriously considering this feature lol).

License

This project is licensed under the MIT License (see for details)

Author

Built by Aryan Vyahalkar | Portfolio