n8n AI Agent Returns 'fetch failed' Error and Responds Slowly

Below is a list of suggestions to help troubleshoot and resolve this issue in your n8n setup:

1. Check for Network Connectivity Issues

The error “TypeError: fetch failed” indicates that the underlying HTTP request did not complete. Verify that your machine can reach any external services used by your AI Agent. On Windows 10, it’s worthwhile to confirm that your firewall or antivirus software isn’t blocking outgoing connections.

2. Verify API Endpoints and Credentials

If your AI agent connects to an external API (for example, OpenAI or a similar provider), double-check that your API keys, endpoints, and any associated permissions are correct. An incorrect endpoint or authentication failure can also lead to fetch requests timing out or failing.

3. Test the Endpoint Independently

Use a tool like curl or Postman to send a request to the API endpoint from your Windows machine. This can help confirm whether the error is specific to n8n or a broader network or service issue.

4. Check for Rate Limits or Quota Issues

Although you mentioned that no proxy is set up, some external APIs enforce rate limits or require sufficient quota/credits. Verify your API account details and usage to ensure you’re not exceeding any limits that might cause requests to fail or respond slowly.

5. Consider Upgrading n8n

While you’re running n8n version 1.78.1, it’s possible that a newer version contains performance improvements or fixes for connectivity issues. Reviewing the release notes and upgrading (if feasible) might resolve the slow responses and fetch failures.

Summary

The error you're encountering generally points to connectivity or configuration issues rather than to proxy settings. By methodically verifying network access, endpoint configuration, API credentials, and rate limits—and keeping your n8n installation up-to-date—you should be able to narrow down and address the root cause.

If you continue to have issues after these steps, consider enabling more detailed logging in n8n to gather additional insight on the error conditions.