🤖 AI Coding Assistant

Never hit API rate limits again

Seamless AI coding with automatic failover across 15+ free AI providers.

A terminal-based AI coding assistant that automatically switches between multiple AI providers whenever one becomes unavailable or rate-limited.

Zero cost. Zero interruptions. Maximum productivity.

✨ Features

| Feature | Description | | ----------------------- | --------------------------------------------------------- | | 🔄 Auto Failover | Automatically switches providers when rate limits are hit | | 🧠 Persistent Context | Remembers conversations across sessions | | 📦 Large Prompt Support | Handles prompts with thousands of lines | | 🆓 Free to Use | Leverages free tiers from multiple providers | | 🚀 Unlimited Workflow | Continue working even when providers throttle requests | | 💬 Dual Mode | General Chat + Coding Assistant modes | | 🔧 Project Operations | Read files, analyze code, execute commands | | 📊 Smart Routing | Tracks provider health and prioritizes working services | | 🔒 Privacy First | Context stored locally in Markdown files |

🎯 Why This Exists

Most AI coding tools depend on a single provider.

When that provider:

• Hits rate limits
• Goes down
• Becomes slow
• Exceeds free quotas

your workflow stops.

AI Coding Assistant solves this problem by automatically rotating between multiple providers behind the scenes.

Example

User Request
      │
      ▼
Groq ❌ Rate Limited
      │
      ▼
Gemini ❌ Temporary Error
      │
      ▼
Cohere ✅ Success
      │
      ▼
Response Returned

You never see the failover process—it just works.

📦 Installation

Global Installation

npm install -g ai-code-assistant

Run Without Installing

npx ai-code-assistant

Install From Source

git clone https://github.com/PraveenMistry/ai-code-assistant.git

cd ai-code-assistant

npm install

npm link

🚀 Quick Start

1. Run Setup Wizard

ai-code --setup

2. Select Providers

Choose providers such as:

• Groq
• Gemini
• Cohere
• Hugging Face
• OpenRouter

3. Enter API Keys

The wizard will guide you through configuration.

4. Start Coding

ai-code

💻 Usage

General Chat Mode (Default)

💻 > What is the capital of France?

💻 > Explain quantum computing in simple terms

💻 > Tell me a joke

💻 > How do I learn Python?

Coding Mode

Switch modes anytime:

💻 > mode

🎭 Mode switched to: CODING MODE

Examples:

💻 > create a React Button component

💻 > fix the bug in server.js

💻 > add error handling to the API

💻 > write unit tests for auth module

Working With Your Project

💻 > analyze

Analyze project structure and requirements.

💻 > list

List all project files.

💻 > read package.json

Read a specific file.

💻 > context

Show conversation history.

🧩 Commands

General Commands

| Command | Description | | ------- | ----------------------------------- | | help | Show help and examples | | mode | Switch between Chat and Coding mode | | context | View conversation history | | clear | Clear current session | | stats | Show provider performance | | exit | Save session and quit |

Coding Commands

| Command | Description | | ----------- | ------------------------- | | analyze | Analyze project structure | | list | List project files | | read | Read file contents |

🔑 Supported Providers

| # | Provider | Free Tier | Status | | - | ------------ | -------------------- | ------------- | | 1 | Groq | 10K tokens/min | ✅ Fastest | | 2 | Gemini | 60 requests/min | ✅ Recommended | | 3 | Cohere | 100K tokens/month | ✅ Stable | | 4 | Hugging Face | 30K tokens/day | ✅ Good | | 5 | OpenRouter | 20 req/min + credits | ⚠️ Limited | | 6 | Cerebras | Unlimited | ⚠️ API Issues | | 7 | Together AI | Free credits | ⚠️ Limited | | 8 | DeepInfra | 10K tokens/day | ⚠️ Limited | | 9 | Lepton AI | 100 req/day | ⚠️ Limited |

Recommended Setup

For best reliability:

1. Groq
2. Gemini
3. Cohere

⚙️ Configuration

Environment Variables (.env)

# Groq
GROQ_API_KEY=gsk_xxxxxxxxx
GROQ_MODEL=mixtral-8x7b-32768
GROQ_PRIORITY=1
GROQ_ENABLED=true

# Gemini
GEMINI_API_KEY=xxxxxxxxx
GEMINI_MODEL=gemini-1.5-flash
GEMINI_PRIORITY=2
GEMINI_ENABLED=true

# Cohere
COHERE_API_KEY=xxxxxxxxx
COHERE_MODEL=command
COHERE_PRIORITY=3
COHERE_ENABLED=true

# Project
PROJECT_PATH=/path/to/project
TRD_PATH=/path/to/requirements.md

# Context
MAX_RESPONSE_TOKENS=2048
MAX_CONTEXT_CHARS=6000
MAX_HISTORY_MESSAGES=10

Setup Wizard

Run:

ai-code --setup

The wizard will:

• Detect existing configuration
• Allow reset or reuse
• Configure providers
• Add API keys
• Set project paths
• Save preferences

📁 Project Structure

your-project/
│
├── .ai-context/
│   ├── session.md
│   ├── history/
│   └── prompts/
│
├── .env
│
└── source files...

💡 Tips

Get Free API Keys

| Provider | URL | Setup Time | | -------- | ------------------------------------- | ---------- | | Groq | https://console.groq.com/keys | ~1 min | | Gemini | https://aistudio.google.com/apikey | ~1 min | | Cohere | https://dashboard.cohere.com/api-keys | ~2 min |

Check Provider Health

💻 > stats

Clear Context

💻 > clear

🧹 Session cleared!

Useful when switching projects or topics.

🔧 Troubleshooting

No Valid Providers Configured

ai-code --setup

or

npm run setup

Provider Failed

Enable multiple providers:

GROQ_ENABLED=true
GEMINI_ENABLED=true

Keyboard Stops Responding

If a provider hangs:

Ctrl + C

Restart the application.

Disable Failing Provider

OPENROUTER_ENABLED=false

📊 Performance

| Provider | Avg Response Time | Success Rate | | -------- | ----------------- | ------------ | | Groq | ~500ms | 99% | | Gemini | ~1200ms | 98% | | Cohere | ~2000ms | 95% |

🛣️ Roadmap

Completed

• ✅ Auto failover
• ✅ Multi-provider support
• ✅ Interactive CLI
• ✅ Persistent context
• ✅ Dual mode support

Coming Soon

• 🔄 Ollama / Local Models
• 🔌 VS Code Extension
• 🌐 Web Dashboard
• 🎙 Voice Commands
• 📈 Advanced Analytics

🤝 Contributing

git clone https://github.com/PraveenMistry/ai-code-assistant.git

cd ai-code-assistant

npm install

npm run setup

npm start

Pull requests are welcome.

📄 License

MIT License

Feel free to use, modify, distribute, and contribute.

Built with ❤️ for developers who hate rate limits

⚡ Type mode to switch between Chat and Coding Mode ⚡