@fetchmax/core


The HTTP Client with AI Superpowers

Modern, lightweight, and extensible HTTP client built on native Fetch API with powerful plugin system and AI capabilities. Works seamlessly in Node.js, Deno, Bun, and browsers.

- 🤖 AI-Powered - Use LLMs without extra libraries (OpenAI, Anthropic, DeepSeek)
- ⚡ Ultra-Lightweight - 3.7 KB core (gzipped)
- 🔌 14 Official Plugins - Modular architecture, use only what you need
- 🎯 TypeScript First - Perfect type inference and safety
- 🌍 Universal - Browser, Node.js, Deno, Bun, Edge runtimes

Quick Start

``bash
npm install @fetchmax/core
`

`javascript
import { HttpClient } from '@fetchmax/core';

const client = new HttpClient({
baseURL: 'https://api.example.com'
});

const response = await client.get('/users');
console.log(response.data);
`

Available Plugins

$3

Use AI without installing separate AI SDKs:

`bash
npm install @fetchmax/plugin-ai-agent # AI foundation (required for AI plugins)
npm install @fetchmax/plugin-ai-mock # Generate realistic mock data
npm install @fetchmax/plugin-ai-translate # Multi-language translation
npm install @fetchmax/plugin-ai-summarize # Content summarization
npm install @fetchmax/plugin-ai-transform # Custom AI transformations
`

$3

Production-ready HTTP utilities:

`bash
npm install @fetchmax/plugin-retry # Auto-retry with exponential backoff
npm install @fetchmax/plugin-cache # Response caching with TTL
npm install @fetchmax/plugin-interceptors # Request/response/error interceptors
npm install @fetchmax/plugin-timeout # Request timeouts
npm install @fetchmax/plugin-logger # Debug logging
npm install @fetchmax/plugin-dedupe # Prevent duplicate requests
npm install @fetchmax/plugin-rate-limit # Rate limiting with queue
npm install @fetchmax/plugin-transform # Data transformation
npm install @fetchmax/plugin-progress # Upload/download progress
`

Usage Example

`javascript
import { HttpClient } from '@fetchmax/core';
import { retryPlugin } from '@fetchmax/plugin-retry';
import { cachePlugin } from '@fetchmax/plugin-cache';

const client = new HttpClient({ baseURL: 'https://api.example.com' })
.use(retryPlugin({ maxRetries: 3, backoff: 'exponential' }))
.use(cachePlugin({ ttl: 300000 }));

// Requests now have auto-retry and caching
const users = await client.get('/users');
`

AI Example

Use LLMs in production without installing separate AI SDKs:

`javascript
import { HttpClient } from '@fetchmax/core';
import { aiAgentPlugin } from '@fetchmax/plugin-ai-agent';
import { aiTranslatePlugin } from '@fetchmax/plugin-ai-translate';

// Setup AI (no openai or @anthropic-ai/sdk needed!)
const aiAgent = aiAgentPlugin({
provider: 'openai', // or 'anthropic' or 'deepseek'
apiKey: process.env.OPENAI_API_KEY,
model: 'gpt-4o-mini'
}).aiAgent;

// Translate API responses on-the-fly
const client = new HttpClient({ baseURL: 'https://api.example.com' })
.use(aiTranslatePlugin({
aiAgent,
targetLanguages: ['es', 'fr', 'de'], // Spanish, French, German
fields: { include: ['title', 'description'] }
}));

// One request, multiple languages!
const response = await client.get('/articles/123');
// Returns: { title: "Hello", title_es: "Hola", title_fr: "Bonjour", ... }
``

Complete Documentation

For complete documentation, guides, and examples:

📚 View Full Documentation on GitHub

The main documentation includes:
- Detailed plugin guides
- AI provider setup (OpenAI, Anthropic, DeepSeek)
- Production-ready examples
- Error handling patterns
- Plugin development guide
- Testing strategies
- And much more!

Package Size

| Format | Size | Gzipped |
|--------|------|---------|
| ESM | 15 KB | 3.7 KB |
| CJS | 17 KB | 4.1 KB |
| TypeScript Defs | 11 KB | - |
| Total (unpacked) | 53.6 KB | 3.7 KB |

What matters: The gzipped size is only 3.7 KB!

Features

- ✅ All standard HTTP methods (GET, POST, PUT, PATCH, DELETE, HEAD, OPTIONS)
- ✅ Request/response interceptors
- ✅ Automatic retries with backoff
- ✅ Response caching
- ✅ Request timeout
- ✅ Progress tracking
- ✅ Request deduplication
- ✅ Rate limiting
- ✅ AI integration (OpenAI, Anthropic, DeepSeek)
- ✅ TypeScript support with perfect type inference
- ✅ Works everywhere: Browser, Node.js, Deno, Bun
- ✅ 100% test coverage (622+ tests passing)

Browser Support

Works in all modern browsers and runtimes that support native Fetch API:
- Chrome, Firefox, Safari, Edge (all recent versions)
- Node.js 18+
- Deno (latest)
- Bun (latest)
- Cloudflare Workers, Vercel Edge Functions, etc.

License

MIT © FetchMax Contributors

Links

- 📦 npm Organization
- 📚 GitHub Repository
- 📖 Full Documentation
- 🐛 Report Issues

---

Made with ❤️ by the FetchMax community