Making our Adyen documentation agent-ready: Bridging the gap between AI and implementation

A recent 30-day analysis of our traffic logs on docs.adyen.com shows why this shift is necessary. Out of more than 10 million total requests, 1.17 million originated from AI bots.

When we looked closer at those numbers, three distinct realities emerged:

• Automated traffic now represents 11.6% of our total volume. Even when accounting for dual-purpose search indexers, machine-driven retrieval has become a foundational part of our traffic footprint.

• Developers are using AI tools for live, contextual troubleshooting. Within our purely generative AI traffic, user-direct assistant agents account for 78% of the total volume. Because these specific bots fetch web content dynamically to answer active developer prompts, it confirms that engineers are relying on AI assistants to generate Adyen integration logic while they are actively coding.

• Traditional web layouts act as a bottleneck for machine runtimes. Currently, 29% of all AI requests land on our homepage and 12% hit our robots.txt file. Bots are spending significant processing cycles trying to reverse-engineer our site hierarchy through trial and error before they ever locate core documentation like our API explorer (7%) or point-of-sale release notes (6%).

For developers, documentation is the source of truth. But traditional documentation websites are often designed for people reading in their web browsers. As a result, the AI agents that many developers use to generate code, troubleshoot integrations, or architect systems, have difficulties parsing the documentation.

To navigate our documentation site, AI agents had to crawl pages over the internet, traverse chains of references, and often hit rate limits along the way. It created a frustrating experience for developers trying to access our content with an agent.  To improve this experience, Adyen is publishing its documentation as a dedicated markdown repository (including llms.txt) that’s synced with our published docs.adyen.com site. By doing this, we are removing hurdles so AI agents can retrieve higher quality context quicker.

We have changed and improved how our content is shown to ensure your AI tools have the best possible information.

• Full context, content, and metadata: Unlike standard web scraping which can miss context, our agent-ready files include full page content enriched with metadata. 

• Adyen code snippets: Every markdown file includes full code samples across all supported languages. Whether your agent is writing in .NET, Java, Python, TypeScript, Ruby, PHP, or Go, it has immediate access to the correct language and implementation patterns. 

• Optimized indexation: Through better site-level indexing and the implementation of the llms.txt standard, agents no longer need to guess where the right information lives. They can find the exact context they need instantly, without the overhead of web traversal.

            ---
    title: "Manage balance accounts"
    description: "View and manage the balance accounts in your platform."
    url: "https://docs.adyen.com/platforms/manage-balance-accounts"
    source_url: "https://docs.adyen.com/platforms/manage-balance-accounts.md"
    canonical: "https://docs.adyen.com/platforms/manage-balance-accounts"
    last_modified: "2026-04-21T14:23:07+02:00"
    language: "en"
    ---

    # Manage balance accounts

    View and manage the balance accounts in your platform.

    A [balance account](/platforms/account-structure-resources/) is an API resource that contains information about the funds held by your user. To manage the balance accounts use the [Configuration API](https://docs.adyen.com/api-explorer/balanceplatform/latest/overview).

    This page contains instructions for common balance account operations. These operations are part of a larger process which includes [verification](/platforms/verification-overview) and [onboarding steps](/platforms/onboard-users/onboarding-steps).

    ## Create a balance account

    You can create balance accounts manually using the Configuration API. To create a balance account:

    1. Make a POST [/balanceAccounts](https://docs.adyen.com/api-explorer/balanceplatform/latest/post/balanceAccounts) request using the account holder ID you want to attach the balance account to.

      **Create a balance account**

      #### curl

      ```bash
      curl https://balanceplatform-api-test.adyen.com/bcl/v2/balanceAccounts \
      -H 'x-api-key: ADYEN_BALANCE_PLATFORM_API_KEY' \
      -H 'content-type: application/json' \
      -X POST \
      -d '{
        "accountHolderId": "AH00000000000000000000001",
        "description": "AccountHolder Primary Balance Account"
      }'
      ```

      #### Java

      ```java
      // Adyen Java API Library v33.0.0
      import com.adyen.Client;
      import com.adyen.enums.Environment;
      import com.adyen.model.balanceplatform.*;
      import java.time.OffsetDateTime;
      import java.util.*;
      import com.adyen.service.balancePlatform.*;

      Client client = new Client("ADYEN_BALANCE_PLATFORM_API_KEY", Environment.TEST);

      // Create the request object(s)
      BalanceAccountInfo balanceAccountInfo = new BalanceAccountInfo()
       .accountHolderId("AH00000000000000000000001")
       .description("AccountHolder Primary Balance Account");

      // Send the request
      BalanceAccountsApi service = new BalanceAccountsApi(client);
      BalanceAccount response = service.createBalanceAccount(balanceAccountInfo, null);
      ```

      

Our goal is to make our powerful financial technology platform as accessible to an AI agent as it is to a human developer. By providing a repository of clean, structured markdown, we unlock a new tier of productivity.

• Context ready: Feed our entire documentation set into your RAG (Retrieval-Augmented Generation) pipeline, application, or custom AI tool with the Copy to Markdown button.

• Reduced friction: No more rate-limiting or broken links during the AI's discovery phase.

• Better accuracy: With direct access to structured data and multi-language samples, the code generated by your AI tools will be more accurate and ready for production.

This release is part of our commitment to building structured infrastructure for AI-driven development. Standardization is the key to this transition. We are following the same philosophy as our Model Context Protocol (MCP) efforts: standardizing how applications provide context to LLMs to support secure, and efficient integrations.

That’s why we’ve made the entire library available for you to use in your local development environments, custom agents, or internal tooling. This is currently available as a zip.

We are excited to see how this changes the way you build integrations with the Adyen platform, so stay tuned for more updates as we continue to share what we’re doing in the (Agent) Developer Experience space.