> For the complete documentation index, see [llms.txt](https://doc.thordata.com/doc/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.thordata.com/doc/scraping/serp-api/integration/mcp-integration/nvidia-nemo-agent-toolkit-integration.md).

# NVIDIA NeMo Agent Toolkit Integration

Connect NVIDIA NeMo Agent Toolkit to Thordata MCP Server for web search, scraping, structured data tools, and configurable agent workflows.

Integrate Thordata MCP Server with NVIDIA NeMo Agent Toolkit to give AI agents access to real-time SERP data and structured search results.

Connect the NVIDIA NeMo Agent Toolkit to the ThorData MCP server (60+ tools) to provide NeMo agents with web search, scraping, and structured data capabilities.

#### Overview

The NVIDIA NeMo Agent Toolkit is an open-source framework for building, analyzing, and optimizing AI agents and workflows. It features a framework-agnostic architecture, YAML-based configuration, built-in MCP support, and unified monitoring across multi-agent systems.

#### Hosted MCP

{% stepper %}
{% step %}
**Get your API token**

* Go to [SERP API-API Token](https://dashboard.thordata.com/serp-api/get-started) to create an API Token;
* Copy your API Token (format as: `sk_7qBRe***************************`）
  {% endstep %}

{% step %}
**Install the NeMo Agent Toolkit**

```
pip install nvidia-nat[mcp]
```

{% endstep %}

{% step %}
**Create workflow configuration**

Create one named: `thordata-mcp-config.yml`

```
function_groups:
  thordata_web:
    _type: mcp_client
    server:
      transport: streamable-http
      url: "https://mcp.thordata.com/YOUR_thordata_API_TOKEN/mcp"
    tool_call_timeout: 120  # Increase timeout for web requests
    auth_flow_timeout: 300
    reconnect_enabled: true
    reconnect_max_attempts: 3


llms:
  nim_llm:
    _type: nim
    model_name: "meta/llama-3.1-8b-instruct"
    temperature: 0.0
    api_key: "${env:NVIDIA_API_KEY}"

workflow:
  _type: react_agent
  tool_names:
    - thordata_web
  llm_name: nim_llm
  max_iterations: 10
  verbose: true
```

{% hint style="info" %}
Replace `YOUR_thordata_API_TOKEN` with the actual token you obtained in the first step.
{% endhint %}
{% endstep %}

{% step %}
**Set environment variables**

```
# Set your NVIDIA API key (for NIM or NGC)
export NVIDIA_API_KEY="your-nvidia-api-key"
```

{% endstep %}

{% step %}
**Verify MCP connection**

Testing the connection to the ThorData MCP server:

```
nat mcp client ping --url "https://mcp.thordata.com/YOUR_thordata_API_TOKEN/mcp"
```

Expected output:

```
Successfully connected to MCP server
Server version: 1.0
Available tools: 15
```

{% endstep %}

{% step %}
**List of available tools**

Check the tools provided by ThorData MCP:

```
nat mcp client tool list --url "https://mcp.thordata.com/YOUR_TOKEN/mcp"
```

{% endstep %}

{% step %}
**Run your first agent**

Execute workflow by querying:

```
nat run --config_file talordata-mcp-config.yml \
  --input "Search for the latest NVIDIA GPU releases and extract their specifications"
```

The agent will use ThorData's web scraping tools to search for and extract information.
{% endstep %}
{% endstepper %}