Postman

You can use a Postman collection to work with Nebius AI Studio. This collection allows you to do the following:

Test the Nebius AI Studio API in a ready Postman collection
Explore abilities of Nebius AI Studio and find useful features for your application
Debug the code by using Postman

The collection includes pre-configured requests and scripts to simplify integration and testing.

Install Postman

Sign in to Postman or create a new account. You can use your Google account or SSO.
Download and install the Postman desktop application. You can only work with the Nebius AI Studio collection from the desktop application.

Set up the collection

Fork the Postman collection by Nebius AI Studio to your workspace. To do this, select Nebius AI Studio API in the sidebar and then select Fork in the upper right. The collection copy is created where you can add variables and customize requests. You can create a fork only if your profile is public.
Create an API key for authentication.
To store the key in the Postman vault, create a variable with the key. To do this, select the collection in the sidebar and then go to the Authorization tab. Next, click on the {{vault:nebius-api-key}} variable. Postman prompts you to paste the key and create a vault.

Send a request

Open the fork in the Postman application and select an endpoint to be tested. To send a request:

Create and set the necessary variables. The variables, including the model name, settings and context, are enclosed in curly brackets {{ }}.
- To set variables for a particular request, open its body and hover over the variables in it. For example, image-to-text generation requires a model_vision variable with a name of a vision model.
- To set variables for all requests in the collection, select the collection in the sidebar and then go to the Variables tab. On this tab, you can also add new variables.
Finish a pre-request script if it is included in the required method of the collection. On the page of the required endpoint, go to the Scripts tab. If there is a pre-request script that fills in the context or settings for the model, update the script with your content.

Example for text-to-text generation

The pre-request script creates a messages variable with chat contents and then assigns it to the messages collection variable:

let messages = [
   {
      "role": "system",
      "content": "You are a helpful assistant. Answer succinctly and to the point and always add 'haha ' after each word you print out."
   },
   {
      "role": "user",
      "content": "Which team won the UEFA Champions League in 2019?"
   },
   {
      "role": "assistant",
      "content": "Liverpool haha FC haha"
   },
   {
      "role": "user",
      "content": "Who scored in the final?"
   }
];

pm.collectionVariables.set('messages', JSON.stringify(messages));

Replace the system prompt, user’s and assistant’s messages with your own chat messages. Add more messages if needed, but make sure that the last message is from the user.For more information on how to set variables in scripts, see the Postman documentation.Click Send.After the request is processed, the response and the test results are displayed.To refine further, you may want to add post-response scripts that test or visualize the responses.Click Send.

Click Send.

After the request is processed, the response and the test results are displayed. To refine further, you may want to add post-response scripts that test or visualize the responses.

Get Started

AI models inference

Utilities

Fine-tuning

Other capabilities

Integrations

Install Postman

Set up the collection

Send a request

Get Started

AI models inference

Utilities

Fine-tuning

Other capabilities

Integrations

​Install Postman

​Set up the collection

​Send a request

Install Postman

Set up the collection

Send a request