Skip to main content

​
Overview

Cursor can work with EURI through its OpenAI-compatible provider settings. This lets you use your EURI API key and route requests through EURI’s unified endpoint instead of calling OpenAI directly. Use this setup if you want to:
  • Access EURI models from inside Cursor.
  • Keep a single API gateway for billing and usage tracking.
  • Switch providers through EURI without changing your Cursor workflow.

​
Prerequisites

Before setting up Cursor, make sure you have:
  1. A valid EURI API key from euron.one/euri.
  2. Access to the EURI base URL:
https://api.euron.one/api/v1/euri

​
Configure Cursor

  1. Open Cursor settings.
  2. Go to the Models section.
  3. Under OpenAI, paste your EURI API key into the OpenAI API key field.
  4. Enable Override OpenAI Base URL (when using key).
  5. Set the base URL to:
https://api.euron.one/api/v1/euri
  1. Save the settings.
  2. If your Cursor version supports custom models, add the EURI model ID you want to use.
Common model examples:
  • gpt-4o-mini
  • gemini-2.5-flash
  • claude-sonnet-4-6
  • llama-4-scout-17b-16e-instruct
See Models for the full list of supported EURI model IDs.

​
Verify the integration

After saving the configuration:
  1. Open a new Cursor chat.
  2. Select the model you configured through the OpenAI provider.
  3. Send a simple prompt like:
Write a one-line summary of what EURI is.
If the request succeeds, Cursor is now sending requests through EURI.

​
How it works

Cursor sends OpenAI-compatible requests, and EURI routes them to the selected upstream model provider. That means you can keep using a familiar Cursor workflow while getting access to multiple providers behind one API.

​
Troubleshooting

​
Verification fails in Cursor

Try the following:
  • Make sure your EURI API key is valid and active.
  • Confirm the base URL is exactly https://api.euron.one/api/v1/euri.
  • Ensure at least one OpenAI-compatible model is enabled in Cursor’s model settings.

​
Model not available

  • Double-check that the model ID exactly matches a model from Models.
  • If your Cursor version does not let you add a custom model ID, use one of the supported OpenAI provider model slots available in your current Cursor build.

​
Requests fail after setup

  • Check that your account has sufficient quota or wallet balance for premium models.
  • Try a non-premium model first, such as gpt-4o-mini or gemini-2.5-flash.
  • If Cursor has network compatibility issues with a custom base URL, switch Cursor’s network compatibility mode to HTTP/1.1 and retry.
Use caseRecommended model
Fast everyday coding helpgpt-4o-mini
Large-context coding tasksgemini-2.5-flash
Higher-quality reasoningclaude-sonnet-4-6
Open-source optionllama-4-scout-17b-16e-instruct

​
Next steps