Comprehensive library for building intelligent AI agents
## 🚀 Quick Start
See how easy to use Ailoy through below examples.
### Get your agent just in _a single line of code_
Check out the simplest python example to build your agent with local models.
```sh
pip install ailoy-py
```
```python
import ailoy as ai
# Create an agent with a local model in a single line of code.
agent = ai.Agent(ai.LangModel.new_local_sync("Qwen/Qwen3-8B"))
# Get the response from the agent simply by calling the `run` method.
response = agent.run("Explain quantum computing in one sentence")
print(response.contents[0].text)
```
### Easy to integrate LLM APIs
Here's the simple javascript example with LLM APIs.
```sh
npm install ailoy-node
```
```js
import * as ai from "ailoy-node";
async function main() {
const lm = await ai.LangModel.newStreamAPI(
"OpenAI", // spec
"gpt-5", // modelName
"YOUR_OPENAI_API_KEY" // apiKey
);
const agent = new ai.Agent(lm);
for await (const resp of agent.run("Please give me a short poem about AI")) {
if (resp.message.contents[0].type === "text") {
console.log(resp.message.contents[0].text);
}
}
}
main().catch((err) => {
console.error("Error:", err);
});
```
### Browser-Native AI (WebAssembly)
You can build your agent entirely in the browser using WebAssembly just in a few lines of code.
```sh
npm install ailoy-web
```
```typescript
import * as ai from "ailoy-web";
// Check WebGPU support
const { supported } = await ai.isWebGPUSupported();
// Run AI entirely in the browser - no server needed!
const agent = new ai.Agent(await ai.LangModel.newLocal("Qwen/Qwen3-0.6B"));
```
### Quick-customizable Web Agent UI Template
Just **Clone** to build your own web agent in minutes.
- https://github.com/brekkylab/ailoy-web-ui
## 🔥 Key Features
### Simple Framework and Powerful Features for AI Agents
- No boilerplate, no complex setup
- **Reasoning**: Extend thinking effortlessly
- **Multi-Modal Inputs**: Process both text and images
- **Extensible Tool Calling**: User-defined functions and **Model Context Protocol (MCP)** tools
- **Retrieval-Augmented Generation (RAG)**: Integrates external knowledge bases without boilerplate
### Cross-Platform & Multi-Language APIs
- Provide **Python** and 
**JavaScript** APIs
- Support 
**Windows**, 
**Linux**, and 
**macOS**
- Support Synchronous and Asynchronous APIs
### Support Web-browser Native AI (WebAssembly)
- Run AI entirely in the browser - no server needed!
### Flexible Model Adoption
- Supports both **local AI** execution and **cloud AI** providers
- Effortlessly switch between open-source and AI services
- Minimal software dependencies — deploy anywhere, from **cloud** to **edge**
### Rust-Powered 
- Fast, memory-safe, minimal dependencies
- Best choice for edge computing and low-resource devices
### Documentation & Community
- [Documentation (Eng.)](https://brekkylab.github.io/ailoy/)
- [Documentation (Kor.)](https://brekkylab.github.io/ailoy/ko/)
- [Discord Community](https://discord.gg/27rx3EJy3P) - Join to ask questions, share your projects, and get help.
## Example Projects
| Project | Description |
| -------------------------------------------------- | ------------------------------------ |
| [Gradio Chatbot](./examples/gradio_chatbot) | Web UI chatbot with tool integration |
| [Web Assistant](./examples/web-assistant-ui) | Browser-based AI assistant (WASM) |
| [RAG Electron App](./examples/simple_rag_electron) | Desktop app with document Q&A |
| [MCP Integration](./examples/mcp_examples) | GitHub & Playwright tools via MCP |
## Installation
> [!WARNING]
> Ailoy is under active development. APIs may change with version updates.
### Python
```sh
pip install ailoy-py
```
### Node.js
```sh
npm install ailoy-node
```
### Browser (WebAssembly)
```sh
npm install ailoy-web
```
## Support Specifications
### Supported AI Models
| Type | Provider & Models |
| ----------- | -------------------------------------------------------------------------------------------------------------------------------------------------- |
| Local Model | 
Qwen3 (0.6B, 1.7B, 4B, 8B, 14B, 32B, 30B-A3B) |
| Cloud API | 
OpenAI (GPT) |
| Cloud API | 
Anthropic (Claude) |
| Cloud API | 
Google (Gemini) |
| Cloud API | 
xAI (Grok) |
### Supported Languags
| Language | Version |
| ---------- | ----------------- |
| Python | 3.10+ |
| JavaScript | ES5+, Node.js 20+ |
### Supported Platforms
| Supported Platform | System Requirements (for Local AI) |
| ------------------ | ---------------------------------- |
| Windows | Vulkan 1.4 compatible GPU |
| Linux | Vulkan 1.4 compatible GPU |
| macOS | Apple Silicon with Metal |
| Web Browser | WebGPU with shader-f16 support |