Core Features
Streaming-First
- Built-in SSE support
- Real-time token streaming
- Efficient LLM responses
- Low-latency experience
Adapter-Based
- Modular architecture
- Multiple AI providers
- Consistent interface
- Type-safe implementation
TypeScript-First
- Full TypeScript support
- Type-safe prompts
- Smart autocompletion
- Runtime validation
Production-Ready
- Error handling & retries
- Rate limiting support
- Logging & monitoring
- Secure credentials
Example Applications
Beginner
OpenAI Chat
Modern chat interface with real-time interactions using OpenAI's GPT models
const chat = createClient({
adapter: openaiAdapter(),
defaultOptions: {
model: 'gpt-4'
}
});Intermediate
Basic RAG Implementation
Retrieval-Augmented Generation system with vector similarity search
const rag = createClient({
adapter: openaiAdapter(),
plugins: [ragPlugin()]
});Intermediate
Basic Agent
AI agent implementation with autonomous task execution and tool usage
const agent = createAgent({
client: aiClient,
tools: [searchTool]
});