Prompt Thing

An open-source, multi-model AI chat platform with image generation and web search capabilities

Prompt Thing is a modern alternative to t3.chat that provides seamless access to multiple AI models, image generation, and advanced reasoning capabilities in a beautiful, intuitive interface.

✨ Features

Core Requirements

• ✅ Chat with Various LLMs - Support for multiple language models and providers
• ✅ Authentication & Sync - User authentication with chat history synchronization
• ✅ Browser Friendly - Web-based interface accessible from any browser
• ✅ Easy to Try - Simple setup and deployment process

Bonus Features

• ✅ Attachment Support - Upload and process files (images and PDFs)
• ✅ Image Generation Support - AI-powered image generation capabilities
• ✅ Syntax Highlighting - Beautiful code formatting and highlighting
• ✅ Resumable Streams - Continue generation after page refresh
• ✅ Chat Branching - Create alternative conversation paths
• ✅ Chat Sharing - Share conversations with others
• ✅ Web Search - Integrate real-time web search
• ✅ Bring Your Own Key - Use your own API keys

🚀 Quick Start

Prerequisites

• Node.js 18+ or Bun
• A Convex account for database and real-time features

Installation

1. Clone the repository

   git clone https://github.com/alifarooq9/promptthing
   cd promptthing

2. Install dependencies

   npm install
   # or
   bun install

3. Set up environment variables

   Create a .env.local file in the root directory:

   # Convex (Required)
   NEXT_PUBLIC_CONVEX_URL=your_convex_url_here
   CONVEX_DEPLOY_KEY=your_convex_deploy_key_here
   
   # Redis (Optional - enables resumable streams)
   REDIS_URL=redis://localhost:6379
   
   # AI Providers (Add keys for models marked as "always" available)
   GOOGLE_GENERATIVE_AI_API_KEY=your_google_key_here
   
   # Image Generation
   RUNWARE_API_KEY=your_runware_key_here

4. Start the development server

   npm run dev
   # or
   bun dev

5. Start the Convex dev server

   npm convex dev

6. Open your browser

   Navigate to http://localhost:3000 to start chatting!

🤖 Supported Models

Text/Chat Models

Provider	Models	Reasoning	BYOK Required
Google Gemini	Gemini 2.5 Pro Preview, 2.5 Flash Thinking, 2.5 Flash, 2.0 Flash Thinking, 2.0 Flash, 2.0 Flash Lite, 2.0 Pro	✅ (Some)	Mixed
OpenAI	GPT-4.1 Nano, GPT-4.1, o3, o3 Mini	✅ (o3 series)	✅
Anthropic	Claude 4 Sonnet, Claude 4 Thinking, Claude 4 Opus, Claude 3.7 Sonnet Thinking, Claude 3.7 Sonnet, Claude 3.5 Sonnet	✅ (Thinking/Opus)	✅
OpenRouter	DeepSeek R1	✅	✅

Image Generation Models

Provider	Models	Features
OpenAI	DALL-E 3, GPT Image Gen	High-quality, versatile
Runware	Runware 101@1, 100@1	Fast generation, cost-effective

🔑 BYOK (Bring Your Own Key) - Some models require you to provide your own API keys

🧠 Reasoning Models - Advanced models with enhanced problem-solving capabilities

🛠️ Development

Adding New Models

Text/Chat Models

1. Add model configuration in config/models.ts:

   const modelDefinitions: Record<string, ModelConfig> = {
     "your-model-id": {
       model: "actual-model-name",
       provider: "provider-name",
       modelName: "Display Name",
       canReason: true, // or false
       supportsWebSearch: true, // or false
       icon: "icon-name",
       availableWhen: "always" | "byok",
       canUseTools: true, // or false
       apiKeyEnv: "ENV_VAR_NAME", // only if availableWhen is "always"
     },
     // ...existing models
   };

2. For new providers, also update:

   • Add provider to Provider type in config/models.ts
   • Install SDK: npm install @ai-sdk/provider-name
   • Add case in createModel() function in lib/models.ts
   • Update isProviderSupported() array in lib/models.ts

Image Generation Models

1. Add model configuration in config/models.ts:

   const imageGenModelDefinitions: Record<string, ImageGenModelConfig> = {
     "model-id": {
       model: "actual-model-name",
       provider: "provider-name",
       modelName: "Display Name",
       imageToImage: true, // or false
       icon: "icon-name",
       availableWhen: "always" | "byok",
     },
     // ...existing models
   };

2. Add provider support in createImageGenModel() function in lib/models.ts

Architecture

• Frontend: Next.js 14 with App Router, Tailwind CSS, Radix UI
• Backend: Convex for database, real-time subscriptions, and serverless functions
• AI Integration: AI SDK for unified model access
• Styling: Tailwind CSS and ShadCN UI
• State Management: Zustand for client-side state

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

⭐ Support

If you find this project helpful, please consider giving it a star on GitHub!