How to Build Open Source Agentic AI Toolkit on AWS Graviton(Keep updating)

AWS Graviton processors, designed for cloud-native workloads, deliver significant cost savings, energy efficiency, and performance improvements for AI/ML applications. In this guide, we’ll explore how to harness Graviton’s ARM-based architecture to build a powerful Agentic AI toolkit using open-source frameworks like LangManus (AI automation framework). We’ll walk through setup, optimization tips, and deployment steps to help you create scalable AI agents on AWS.

Why AWS Graviton for Agentic AI?

AWS Graviton processors (e.g., Graviton4) are optimized for modern workloads, offering:

- **Cost Efficiency**: Up to 40% better price-performance vs. comparable x86 instances.

- **Sustainability**: Reduced energy consumption for eco-friendly AI.

- **ARM Compatibility**: Native support for frameworks like PyTorch, TensorFlow, and Python-based AI tools.

For Agentic AI—where autonomous systems require efficient scaling—Graviton’s cost-performance balance makes it ideal.

Setting Up Your AWS Graviton Environment

### **Step 1: Launch a Graviton Instance**

1. Navigate to AWS EC2.

2. Select an ARM-compatible OS (Amazon Linux 2023).

3.Choose a Graviton-based instance (e.g., `c8g` or `m8g`).

### **Step 2: Install Dependencies**

1
2
3
4
sudo dnf install python3-pip nodejs -y
pip3 install --upgrade pip
pip3 install uv
npm install -g pnpm

Installing LangManus API server on Graviton

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
# Clone the repository
git clone https://github.com/langmanus/langmanus.git
cd langmanus

# Create and activate virtual environment through uv
uv python install 3.12
uv venv --python 3.12

source .venv/bin/activate  

# Install dependencies
uv sync

# Configure environment
cp .env.example .env
# Edit .env with your API keys

# Run the project
uv run server.py

Now the LangManus API server is launching

Image not found

Installing LangManus Web UI on Graviton

1
2
3
# Clone the repository
git clone https://github.com/langmanus/langmanus-web.git
cd langmanus-web

Create a .env file in the project root and configure the following environment variables:

NEXT_PUBLIC_API_URL: The URL of the LangManus API server.

1
2
3
4
5
# Install dependencies
pnpm install

# Run the project in development mode
pnpm dev

Now the LangManus Web UI is launching

Image not found

## **5. Deployment and Scaling**

- **Containerize**: Package your toolkit in Docker with ARM-based images.

- **Orchestrate**: Deploy on AWS EKS/ECS/Fargate (ARM-compatible) for scaling.

- **Monitor**: Use Amazon CloudWatch to track performance and costs.

Conclusion

By combining AWS Graviton’s efficiency with open-source tools like LangManus, you can build high-performance Agentic AI systems at a fraction of the cost. Start with the approach to maximize savings while ensuring stability.

Ready to dive deeper?

- Explore [LangManus Documentation](https://github.com/langmanus/langmanus)

- Learn more about [AWS Graviton](https://aws.amazon.com/graviton)

Any opinions in this post are those of the individual author and may not reflect the opinions of AWS.

Select your cookie preferences

Site Terms, Privacy, and more.