Balancing data for agents

An MCP server that brings balancing market data from 25+ sources to an agent you control

#Use cases

Analyzing Finnish mFRR balancing energy prices in Claude Cowork

Exporting France aFRR energy prices into an Excel sheet in Claude Cowork

Estimating mFRR down capacity revenue potential for a 50 MW solar park in Estonia

#Getting started

Get started with a 7-day free trial by setting up the MCP server in one of the following ways:

#
1

Open (or create) Antigravity's MCP config file:

  • macOS / Linux: ~/.gemini/config/mcp_config.json
  • Windows: C:\Users\<USERNAME>\.gemini\config\mcp_config.json
2

Add the balancing-services entry inside mcpServers and save:

{
  "mcpServers": {
    "balancing-services": {
      "serverUrl": "https://agents.balancing.services/mcp"
    }
  }
}

Antigravity uses serverUrl (not url) for remote servers. If the file already has a mcpServers block, add balancing-services inside it instead of replacing the file.

3

In the Antigravity app, open Settings, go to Customizations, and click Refresh under Installed MCP Servers. On first tool use, Antigravity opens your browser for the OAuth login flow — once authorized, the balancing-services tools are available.

In the Antigravity CLI, run /mcp and start authentication for balancing-services from there.

Custom MCP servers are supported in the Antigravity app and CLI, not the Gemini web app. For more on connecting MCP servers, see the official MCP documentation.

#

Follow OpenAI's official guide: Developer mode and MCP apps in ChatGPT.

Adding custom MCP apps currently requires Developer mode, which is in beta. On Business, Enterprise, and Edu plans, a workspace admin must enable it before individual users can turn it on. Plus and Pro users can self-enable it.

When you reach the New App form, paste the URL below into the MCP Server URL field and leave Authentication set to OAuth:

https://agents.balancing.services/mcp
#

In the Claude web, desktop, or Cowork app:

  • Go to Customize > Connectors (on the web, open it directly).
  • Click +, then Add custom connector.
  • Give it a name (e.g. Balancing Services), paste the URL below as the remote MCP server URL, and leave Advanced settings empty.
https://agents.balancing.services/mcp
  • Click Connect to authenticate — the server uses OAuth and triggers the login flow automatically.

Tip: query results are served as CSV files from agents.balancing.services. If you add that domain to Claude's network access allowlist, the agent can download results directly instead of pulling the rows into its context — drastically reducing token consumption on larger queries.

Custom connectors are available on Free, Pro, Max, Team, and Enterprise plans (on Team and Enterprise, an organization owner adds the connector). For full details and plan-specific steps, see Anthropic's guide: Get started with custom connectors using remote MCP.

#
1

In a terminal, run:

claude mcp add --transport http balancing-services https://agents.balancing.services/mcp

Using the VS Code extension? Open VS Code's integrated terminal (Terminal → New Terminal) and run the same command there — the extension reads the same configuration.

2

Start a new session.

In the VS Code extension, open a new conversation with the Claude Code icon in the Activity Bar.

3

Type /mcp and follow the authentication flow for balancing-services.

For more on managing MCP servers in Claude Code, see the official MCP documentation.

#
1

In a terminal, run the command below. Codex triggers the OAuth login flow automatically as it adds the server.

codex mcp add balancing-services --url https://agents.balancing.services/mcp

If the login flow doesn't start automatically, run codex mcp login balancing-services from a terminal to start it manually.

2

Start a new session — the balancing-services tools are now available.

For more on connecting MCP servers to Codex, see the official MCP documentation.

#
1

Open Cursor Settings.

2

Select Tools & MCPs.

3

Click Add custom MCP and add the Balancing Services entry inside mcpServers:

    "Balancing Services": {
      "type": "http",
      "url": "https://agents.balancing.services/mcp"
    }
4

Click Connect on Balancing Services and follow the authorization flow.

For more on installing MCP servers in Cursor, see the official MCP documentation.

#

Microsoft Copilot connects to remote MCP servers through Copilot Studio, where the server is added as a tool on an agent. The standard Microsoft 365 Copilot chat can't add a remote MCP server on its own. Adding MCP tools requires a Copilot Studio license and an environment where you can build or edit agents.

1

In Copilot Studio, open the agent you want to extend (or create a new one).

2

Go to the agent's Tools page, click Add a toolNew toolModel Context Protocol.

3

In the wizard, give it a name (e.g. Balancing Services), paste the URL below as the Server URL, and set Authentication to OAuth 2.0:

https://agents.balancing.services/mcp

The server supports Dynamic Client Registration with discovery, so Copilot Studio configures the OAuth connection automatically — there's no client ID or secret to enter.

4

Create the tool, then trigger it in a test chat with the agent. On first use you're prompted to sign in and authorize — once you complete the OAuth flow, the balancing-services tools are available.

For full details and plan requirements, see Microsoft's guide: Connect your agent to an existing MCP server.

#

Copy the instructions below and give them to your agent, or follow them manually.

Add the Balancing Services MCP server to your MCP configuration:

{
  "mcpServers": {
    "balancing-services-api": {
      "type": "http",
      "url": "https://agents.balancing.services/mcp"
    }
  }
}

The server uses OAuth — most MCP clients trigger the auth flow automatically on first tool use.

#Pricing

Individual Subscription

€59/month
  • Full access to all European balancing market data
  • Tied to a specific user
  • Cancel anytime

Custom

Tailored to your needs

  • Bulk purchases
  • Data access for fully autonomous agents
  • Other use cases
Contact Us

Want to try it first? For a 7-day free trial, follow the "Getting started" section above.