Overview

The easiest way to build AI-powered applications, and autonomous agents on the JVM. Connect to OpenAI, Claude, Gemini, and more with under 10 lines of code.

BoxLang AI is the fastest path from idea to production-grade AI applications. With a single line of code, connect to 12+ providers including OpenAI, Claude, Gemini, Grok, and Ollama. BoxLang AI delivers a unified API, autonomous agent architecture, multi-tenant memory systems, and enterprise-ready integrations—giving you everything needed to build sophisticated AI applications without the complexity.

We recommend you use simple chat functions (aiChat(), aiMessage()) when you want to quickly add AI capabilities to existing applications. Use AI Agents and Pipelines when you have more advanced needs that require autonomous reasoning, tool calling, memory systems, RAG, and multi-step workflows.

BoxLang AI agents are built with composable pipelines and memory systems to provide streaming execution, tool integration, multi-tenant isolation, persistence, and more. You can start simple and scale to complex workflows as your needs grow.

📚 In This Section

📦 Installation

Quick guide to installing BoxLang AI.

What you'll learn:

• Installing via BoxLang Module Installer, CommandBox, or package dependencies

• Basic configuration setup

• Running Ollama with Docker for production deployments

• Verification and next steps

Time: 5 minutes

---

🧩 Provider Setup & Configuration

Comprehensive guide to configuring all supported AI providers.

What you'll learn:

• Provider comparison and recommendations

• Getting API keys for 12+ cloud providers (OpenAI, Claude, Gemini, etc.)

• Setting up Ollama for local AI (no API costs!)

• Configuration best practices

• Environment variables and security

• Multiple provider management

• Troubleshooting provider issues

Time: 10-15 minutes

• 🤖 Multi-Provider Support - OpenAI, Claude, Gemini, Ollama, Grok, Groq, DeepSeek, Perplexity, and more

• 💬 Simple Chat Interface - Start with one-line AI conversations

• 🔄 Composable Pipelines - Build complex AI workflows by chaining operations

• 🧠 Intelligent Agents - Create autonomous agents with memory and tools

• 📊 Structured Output - Extract data into classes, structs, or arrays

• 🎙️ Multimodal Content - Process images, audio, video, and documents

• 🛠️ Real-Time Tools - Enable AI to call functions and APIs

• 💭 Memory Systems - Maintain conversation context across interactions

• 📡 Streaming Support - Real-time response streaming for better UX

🧭 Quick Navigation

🆕 New to BoxLang AI?

Installation Guide Get the module installed in minutes.

Provider Setup Configure your AI providers.

Quickstart Tutorial Your first AI conversation in 5 lines of code.

---

⚡ Quick Start Guide

Your first AI conversation in 5 lines of code, plus essential patterns and examples.

What you'll learn:

• Making your first AI chat request

• Understanding basic BIF usage (aiChat, aiMessage, aiModel)

• Provider switching and model selection

• Streaming responses in real-time

• Working with structured output

• Building your first AI agent with tools and memory

• Common patterns and best practices

Time: 15-20 minutes

---

📖 Key Concepts

Essential AI terminology and concepts explained in plain language.

What you'll learn:

• AI & Machine Learning fundamentals (LLMs, training vs inference)

• Language model parameters (temperature, top-p, tokens, context windows)

• Messages & conversations (roles, system messages, multi-turn)

• Embeddings & vectors (dimensions, cosine similarity, vector databases)

• Memory systems (conversation, vector, hybrid, multi-tenant)

• RAG (Retrieval Augmented Generation) workflow

• Tools & function calling

• Streaming & async operations

• Pipelines & composition

• Providers & services comparison

• Tokens & costs

Time: 30 minutes reference

---

❓ Frequently Asked Questions

Quick answers to the most common questions about BoxLang AI.

Topics covered:

• Getting started (why use BoxLang AI, free options, offline usage)

• Providers & models (comparison, recommendations, model selection)

• Costs & pricing (optimization, estimation, budget management)

• Performance & reliability (response variation, speed, handling downtime)

• Features & capabilities (structured output, tools, memory, RAG, multimodal)

• Memory & context (conversation vs vector, multi-tenant isolation)

• Security & privacy (data privacy, prompt injection, API keys)

• Troubleshooting (common errors, rate limits, context length)

• Best practices (aiChat vs aiAgent, memory sizing, caching, error handling)

Time: Quick reference

---

🎯 Learning Path

We recommend this progression:

1. 📦 Install - Get the module installed (5 min)

2. 🧩 Configure Providers - Set up your AI providers (10 min)

3. ⚡ Quick Start - Your first AI conversation (10 min)

4. 📖 Key Concepts - Learn AI terminology (reference as needed)

5. 💬 Basic Chatting - Simple interactions and parameters (15 min)

6. 🎯 Advanced Chatting - Streaming, tools, multimodal content (20 min)

7. 🤖 AI Agents - Build autonomous agents (30 min)

💡 Pro Tip: Keep the FAQ handy for quick answers to common questions!

---

💡 Quick Examples

Simple Chat

result = aiChat( "What is BoxLang?" );
println( result );

Structured Output

person = aiChat(
    messages: "Extract info: John is 30 and lives in NYC",
    options: {
        returnFormat: {
            name: "string",
            age: "numeric",
            city: "string"
        }
    }
)
println( person.name ) // "John"

AI Agent

// Create an agent with defaults:
// - memory from global config (window) by default
// - model and provider from global config
agent = aiAgent( name: "MyAgent" ).withInstructions( "You are a helpful assistant" );

// Run the agent with a user prompt
response = agent.run( "Hello! Remember my name is Alice." );

---

🔗 Related Documentation

After mastering the basics, explore these advanced topics:

• 🔄 Pipelines - Build composable AI workflows

• 💭 Memory Systems - Maintain conversation context

• 🛠️ Tools - Enable AI to call functions

• 📄 Document Loaders - Process various file formats

• 🔮 Vector Memory - Semantic search with embeddings

---

❓ Need Help?

• 🐛 Found a bug? Report it on GitHub

• 💡 Have an idea? Start a discussion

• 🆘 Professional Support? AI Services