A modern web application for batch processing text completions through Anthropic's Claude AI model. This application allows you to efficiently create, manage, and monitor batches of AI completion requests.

• Batch Processing: Upload, configure, and process large batches of prompts with Claude
• Dashboard Interface: Monitor real-time progress of your batch completions
• Authentication: Secure user authentication with email/password and OAuth providers
• API Key Management: Create and manage API keys for programmatic access
• User Roles: Administrative controls and user role management
• Robust Security: Comprehensive security headers and authentication flows
• Responsive Design: Modern UI built with React, Tailwind CSS, and Shadcn UI components

• Framework: Next.js 15 with App Router and React Server Components
• Authentication: NextAuth.js with database adapter
• Database: PostgreSQL with Prisma ORM
• Schema Enhancement: ZenStack for access policies
• Form Management: React Hook Form with Zod validation
• Styling: Tailwind CSS with Shadcn UI components
• State Management: Zustand and React Query
• UI Components: Radix UI primitives
• AI Integration: Anthropic API for Claude AI access

• Node.js 18+ and npm/yarn/pnpm
• PostgreSQL database
• Anthropic API key

1. Clone the repository:

   git clone https://github.com/yourusername/claude-batch.git
   cd claude-batch

2. Install dependencies:

   npm install
   # or
   yarn install
   # or
   pnpm install

3. Set up environment variables:

   • Copy .env.example to .env.local
   • Fill in the required environment variables:

     # Database
     DATABASE_URL="postgresql://postgres:password@localhost:5432/claude_batch"
     
     # Anthropic API
     ANTHROPIC_API_KEY="your-anthropic-api-key"
     
     # NextAuth
     NEXTAUTH_SECRET="your-nextauth-secret"
     NEXTAUTH_URL="http://localhost:3000"
     
     # Optional OAuth Providers
     # GITHUB_ID="your-github-client-id"
     # GITHUB_SECRET="your-github-client-secret"
     # GOOGLE_ID="your-google-client-id"
     # GOOGLE_SECRET="your-google-client-secret"
     

4. Initialize and migrate the database:

   npx prisma migrate dev

5. Start the development server:

   npm run dev
   # or
   yarn dev
   # or
   pnpm dev

6. Open http://localhost:3000 in your browser to see the application.

The application uses the following core data models:

• User: User accounts with authentication information
• Batch: Represents a collection of completions to be processed
• Completion: Individual AI completion requests within a batch
• ApiKey: API keys for programmatic access to the application

• Email/Password authentication
• OAuth providers support (GitHub, Google)
• Session-based authentication with JWT
• Protected routes via middleware
• API key authentication for programmatic access

The application provides both a web interface and API endpoints for managing batches:

1. Create an API key in the settings
2. Use the key to authenticate API requests:

   curl -X POST https://your-deployment-url/api/batches \
     -H "Authorization: Bearer YOUR_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{
       "name": "My Batch",
       "prompts": ["Tell me about AI", "Explain quantum computing"]
     }'

This application supports the following Claude models:

• Claude 4 (Latest Generation):

  • claude-opus-4-20250514 - Our most capable and intelligent model with 32K max output tokens
  • claude-sonnet-4-20250514 - High-performance model with 64K max output tokens
  • Features: 200K context window, extended thinking support, vision capabilities, March 2025 training cutoff

• Claude 3.7:

  • claude-3-7-sonnet-20250219 - High-performance model with early extended thinking (up to 128K output with beta header)
  • Features: 200K context window, extended thinking support, vision capabilities, October 2024 training cutoff

• Claude 3.5:

  • claude-3-5-sonnet-20241022 - Latest Claude 3.5 Sonnet (October 2024)
  • claude-3-5-haiku-20241022 - Fastest model for quick responses
  • claude-3-5-sonnet-20240620 - Previous Claude 3.5 Sonnet (June 2024)
  • Features: 200K context window, 8K max output tokens, vision capabilities

• Claude 3 (Legacy):

  • claude-3-opus-20240229 - Most capable Claude 3 model
  • claude-3-sonnet-20240229 - Balanced Claude 3 model
  • claude-3-haiku-20240307 - Fastest Claude 3 model
  • Features: 200K context window, 4K max output tokens, vision capabilities

To verify model availability or test your Anthropic API connection, you can use the provided scripts:

# Check available models (JavaScript)
cd scripts && node check-models.js

# Check available models (TypeScript)
cd scripts && npx ts-node check-models.ts

# Test batch API functionality (JavaScript)
cd scripts && node check-batch-api.js

# Test batch API functionality (TypeScript)
cd scripts && npx ts-node check-batch-api.ts

The scripts will automatically use the Anthropic API key from the main project's .env.local file, so you don't need to create a separate .env file in the scripts directory.

• Batch Structure: Each batch request must include a list of individual message requests, each with a unique custom_id and message parameters.
• Propagation Delay: After creating a batch, there might be a short delay (5-15 seconds) before it's fully available in Anthropic's system.
• Model Name Validation: Ensure you're using a valid model name as shown above. Invalid model names will result in a not_found_error.
• Error Handling: The application implements retry logic for handling temporary availability issues with newly created batches.

1. Create a Batch: Upload a CSV or JSON file with prompts or enter prompts manually
2. Configure: Set Claude model, temperature, and other parameters
3. Process: Start the batch and monitor progress in real-time
4. Results: Download or export results when processing completes

claude-batch/
├── app/                   # Next.js App Router
│   ├── (auth)/            # Authentication pages
│   ├── (dashboard)/       # Protected dashboard pages
│   ├── api/               # API routes
│   └── page.tsx           # Landing page
├── components/            # Reusable React components
├── config/                # Configuration files
├── hooks/                 # Custom React hooks
├── lib/                   # Utility functions and services
├── prisma/                # Database schema and migrations
│   ├── schema.prisma      # Prisma schema
│   └── zenstack.schema    # ZenStack enhanced schema
├── public/                # Static assets
└── types/                 # TypeScript type definitions

• npm run dev: Start development server
• npm run build: Build for production
• npm run start: Start production server
• npm run lint: Run ESLint

This project is licensed under the MIT License.