No VPN
Workflow requests go from your n8n server to api.chatqt.com — without needing a tunnel to openai.com.
Developers
n8n with the ChatQT API.
In n8n workflows, use the OpenAI node with a custom credential — enter the ChatQT API key and Base URL and call GPT, Claude and other models without a VPN from Iran.
OpenAI credential
API Key: YOUR_CHATQT_API_KEY Base URL: https://api.chatqt.com/api/v1
Why ChatQT?
n8n automation without connection hassle.
Instead of a direct OpenAI account, which is often unavailable from Iran, configure the same OpenAI node with ChatQT's compatible endpoint — dozens of models from one credential.
Stable access
VPN drops no longer halt your production workflows.
Transparent payment and usage
API key from developer console; cost is based on usage — details on the Pricing.
Prerequisites
Be ready before you start.
- 1 An instance of n8n (Cloud or self-hosted) with access to Credentials.
- 2 account ChatQT and an API key from the console.
- 3 Basic familiarity with building workflows in n8n (Trigger + nodes). Official docs: docs.n8n.io
Step 1
Create an OpenAI-type credential
In n8n, create a credential of type OpenAI API . Instead of the OpenAI key, enter the ChatQT key and override the base URL.
- In n8n, go to Credentials → Add credential.
- the type OpenAI API and select it.
- In the field API Key Copy the key from the ChatQT console and paste it.
- the field Base URL (in some versions Host or Override URL) set it to the address below.
- Save and test the credential in a workflow.
Credential values
API Key: YOUR_CHATQT_API_KEY Base URL: https://api.chatqt.com/api/v1
For the OpenAI SDK and most compatible clients, the same path is /api/v1 — details on the API documentation.
Step 2
Add the OpenAI node to a workflow
After creating the credential, add the OpenAI node and select the model by its console ID.
- A new or existing workflow → + → search OpenAI.
- Set Resource to Message a model (or Chat / Text).
- Select the credential you created in the previous step.
- In the field Model Enter the exact ID from the console (e.g.,
openai/gpt-4.1). - Fill in Prompt / Messages and save the workflow.
Simple example: Manual Trigger → OpenAI (Message a model) → show output. To start, test a fixed message in the node.
Popular models
| Model | ID in the node |
|---|---|
| GPT 4.1 | openai/gpt-4.1 |
| Claude Sonnet 4.6 | anthropic/claude-sonnet-4.6 |
| DeepSeek V4 | deepseek/deepseek-v4 |
The full, up-to-date model list is in the console.chatqt.com .
Step 3
Test the workflow
- Click Test workflow or manually run the node.
- The output should show the model's response; check usage in the ChatQT console.
401 error (Unauthorized)
Check the API key in the credential; it should be copied from the ChatQT console with no extra spaces.
Model not found
Copy the model ID exactly from the console (e.g., openai/gpt-4.1) — the display name is not the same as the API slug.
Connection won't establish
The Base URL should be https://api.chatqt.com/api/v1 (with /v1 at the end).
FAQ
Common questions
Why the OpenAI node and not a separate ChatQT node?
The ChatQT API is OpenAI-format compatible; by overriding the Base URL, the standard n8n node is enough and there's no need to install a community node.
Where is the model list?
All usable IDs are shown in the developer console — write that same slug in the Model field.
What's the difference between /api and /api/v1?
For n8n and the OpenAI SDK, use
/api/v1 . Claude CLI uses the path without v1 — Claude CLI guide and API documentation.How is the cost calculated?
Like other API requests — pay-as-you-go. The rate table is on the pricing page.
Ready to automate with n8n?
Get an API key, set up the OpenAI credential and run your first workflow.