🔥20+ TypeScript AI Agent Examples with Source Codes

Why I Put These Examples Together

Simple: I think they're useful! If you're curious about AI agents – whether you're a beginner or have some experience, these examples offer a practical look.

Instead of just theory, you get to see how agents are built with straightforward approaches. Our hope is that they help you understand how agents operate and maybe even spark an idea for your own project.

The Real world AI Agent examples

We built VoltAgent, a lightweight, open-source TypeScript framework designed to make building AI agents much, much easier.

Here's a tour of what you can build. We've put some of the real juicy ones upfront.

Find all the code here: VoltAgent Examples on GitHub

📄 Google Drive MCP Integration (with-google-drive-mcp)

Ever wanted your AI to rummage through your Google Drive? This shows you how, using the magic of Model Context Protocol (MCP).

👉 Get this example source code on Github

🧠 RAG Chatbot (with-rag-chatbot)

Build a smarty-pants chatbot! This one uses Retrieval Augmented Generation (RAG) to answer questions based on its own knowledge stash.

👉 Get this example source code on Github

📊 Peaka MCP Integration (with-peaka-mcp)

Description: Make your agent data-savvy by connecting it to Peaka. Query databases, spreadsheets, you name it!

👉 Get this example source code on Github

🤖 GitHub Repo Analyzer (Multi-Agent) (github-repo-analyzer)

This one's a bit of a show-off. A multi-agent setup that dissects GitHub repos. See agents working together – it's like a digital beehive!

👉 Get this example source code on Github

💾 Supabase Integration (with-supabase)

Hook your agent up with Supabase for a solid backend. Think storing chat histories or user profiles.

👉 Get this example source code on Github

🕊️ Turso Integration (with-turso)

Get your agent talking to Turso, the distributed SQLite database that's pretty fly.

👉 Get this example source code on Github

💡 Xsai Integration (with-xsai)

An example showing how to get VoltAgent playing nicely with Xsai.

👉 Get this example source code on Github

🚀 Vercel AI SDK (with-vercel-ai)

Deploy your agent to the edge! This example uses the Vercel AI SDK for that serverless goodness.

👉 Get this example source code on Github

🌐 Next.js Frontend (with-nextjs)

Go full-stack! VoltAgent for the brains, Next.js for the beauty. Build a complete AI app.

👉 Get this example source code on Github

☁️ Google Vertex AI (with-google-vertex-ai)

Tap into the power of Google Cloud's Vertex AI models with your VoltAgent.

👉 Get this example source code on Github

🇬 Google AI (Gemini) (with-google-ai)

Use Google's shiny Gemini models in your VoltAgent creations.

👉 Get this example source code on Github

🧱 Base Example (base)

The OG. A super minimal VoltAgent setup to get you off the ground, like, now.

👉 Get this example source code on Github

---

🛠️ More Specialized Examples

🔗 Composio MCP Integration (with-composio-mcp)

Connect to a whole universe of tools and platforms using Composio's MCP connector.

👉 Get this example source code on Github

⚡ Groq AI (with-groq-ai)

Need speed? Leverage Groq AI's LPU inference engine for some seriously fast LLM responses.

👉 Get this example source code on Github

🤝 Generic MCP (with-mcp)

A plain vanilla example of using the Model Context Protocol (MCP) for tool integration. Good for understanding the basics.

👉 Get this example source code on Github

🎭 Playwright Browser Automation (with-playwright)

Let your agent surf the web! Well, automate browser tasks using Playwright, which is almost the same thing.

👉 Get this example source code on Github

🔍 Custom Retrieval (with-retrieval)

Go beyond basic RAG and implement your own fancy retrieval logic for your agent.

👉 Get this example source code on Github

🧩 Sub-Agents (with-subagents)

Agentception! Explore how to create and manage sub-agents within a main VoltAgent.

👉 Get this example source code on Github

🤔 Thinking Tool (with-thinking-tool)

Give your agent a moment to ponder. This example uses a "thinking" tool for complex reasoning.

👉 Get this example source code on Github

🔧 Custom Tools (with-tools)

Shows how to give your agent its own special skills by defining and integrating custom tools.

👉 Get this example source code on Github

🗣️ Voice with ElevenLabs (with-voice-elevenlabs)

Make your agent chatty with super realistic voice output using ElevenLabs.

👉 Get this example source code on Github

🎙️ Voice with OpenAI (with-voice-openai)

Or, use OpenAI's voice models to give your agent a voice (and ears!).

👉 Get this example source code on Github

🔌 Zapier MCP Integration (with-zapier-mcp)

Connect your agent to basically everything (thousands of apps!) through Zapier using MCP.

👉 Get this example source code on Github

---

So, You Want to Build AI Agents Without Tearing Your Hair Out?

Building AI in JS often meant choosing between painful boilerplate or restrictive No-Code tools, and debugging felt like staring into a black box.

So we built VoltAgent, a lightweight, open-source TypeScript framework designed to make building AI agents much, much easier.

• It's open source and community-driven.The repo: https://github.com/VoltAgent/voltagent
• It provides helpful structure, giving you core building blocks for agents (tools, memory, state) so you're not reinventing the wheel.
• It has a built-in visual debugging console that connects locally to your running agent (no data leaves your machine!). You can visually trace execution step-by-step, inspect messages, and see the flow – think n8n-style but for agents. Seriously, it's a lifesaver.

Got questions or built something cool?

• Join us Discord community

• Follow us on X

Dive in, get inspired, and maybe even contribute your own! 😉

What Now?

Hopefully, these examples give you a solid launchpad. VoltAgent is all about making your life easier when building these AI marvels. So, grab an example, tweak it, break it, make it your own.