avatarValentin Despa

Summary

The provided content is a comprehensive guide on how to get started with the OpenAI API using Postman, detailing the process of obtaining an API key, setting up requests, and utilizing the Completions API with practical examples and explanations of key concepts.

Abstract

The article "Getting Started with the OpenAI API with Postman" offers a step-by-step tutorial on integrating OpenAI's powerful AI tools into applications via their API. It begins by instructing readers on how to create an OpenAI account and generate an API key, which is essential for accessing the API. The guide then transitions into using Postman, a popular API development tool, to send requests to the OpenAI API, demonstrating how to list available models and configure authentication with the API key. Furthermore, the article explains how to use the Completions API to generate text with the text-davinci-003 model, providing a JSON payload example and discussing the significance of parameters like model, prompt, max_tokens, and temperature. The author also clarifies OpenAI's pricing model, emphasizing that while new users receive initial credits, continued use of the API is based on a pay-as-you-go system. The article concludes with an invitation for readers to explore more API endpoints and share their thoughts or requests for future tutorial topics.

Opinions

  • The author believes that creating well-researched and concise content is valuable and encourages readers to support the work by engaging with the content (commenting, sharing, and clapping).
  • The author suggests that learning how to use Postman for API testing is beneficial, promoting an online course for those interested in gaining a deeper understanding of API testing fundamentals.
  • The author expresses that the text-davinci-003 model is one of the most advanced language generation models available, capable of producing coherent and expressive text across various applications.
  • The author emphasizes the importance of understanding the cost associated with using the OpenAI API, highlighting the concept of tokens as a unit of measurement for both input and output, which affects pricing.
  • The author provides a practical example to demystify the concept of tokens, relating them to the number of words and characters to give readers a clearer understanding of the API's cost structure.
  • The author clarifies the role of the temperature parameter in generating AI responses, explaining that it influences the randomness and creativity of the output.
  • The author indicates that while OpenAI offers an initial credit to new users, the service operates on a paid basis, with pricing details provided to ensure users are well-informed about the costs associated with API usage.

Getting Started with the OpenAI API with Postman

OpenAI has been in the headlines since it launched tools such as DALL·E 2, Jukebox, and ChatGPT. OpenAI tools have a web interface but also an API that can be used to integrate these tools into your own applications. In this blog post, we will look at how to use the OpenAI API in Postman, a popular API development and testing tool.

Last update: March 2023

Getting an API key for OpenAI

OpenAI is a research laboratory focusing on creating and promoting friendly artificial intelligence. The company has an API (Application Programming Interface) that allows developers to access and utilize the power of its AI models in their own applications.

To create an OpenAI API key, you will need to have an OpenAI account. Here is the process for creating an OpenAI API key:

Step 1. Go to the OpenAI website and sign up for an account: https://platform.openai.com/

Step 2. After you have created an account, log in to your OpenAI account.

Step 3. In the top right corner of the screen, click on your account name and then select “API Keys” from the drop-down menu.

Step 4. On the API Keys page, click the “Create new secret key” button.

Step 5. The API key will be displayed on the screen. Make sure to copy the key and store it in a secure location, as it will not be shown again. You can also leave this page open for now.

Once you have an API key, you can start using the API in Postman.

Accessing the OpenAI API from Postman

To ensure everything is working fine, let’s first run a very simple request from Postman: let’s try and get a list of all available models in OpenAI.

Step 1. Open Postman and create a new request. You can do this by clicking on the “New” button in the top left corner of the Postman window.

Step 2. In the “New Request” dialog, make sure the “GET” method is selected and enter the following URL:

https://api.openai.com/v1/models

Step 3. To save this request in a collection in Postman, click the “Save” button in the top right corner of the request window.

In the “Save Request” window that appears, click on “New Collection”.

Create a new collection named “OpenAI”. You can enter a name for the collection in the field that appears.

Name the request “List models”. When you are finished, click the “Save” button to save the request to the selected collection.

The request will now be saved in the selected collection and can be accessed from the “Collections” tab on the left side of the Postman window.

Step 4. Locate the collection, open the context menu and click on “Edit”.

Step 5. Click on the “Authorization” tab, and from Type, select “Bearer Token”.

In the “Token” field, paste the OpenAI API key you have previously generated.

Step 6. Select the token value, and you should see the dialog “Set as variable” appearing.

Select “Set as a new variable” and name the variable openAIToken. Select the “Collection” scope and click on “Set variable”.

Finally, save all changes in the collection by clicking on “Save”.

Step 7. Go back to the “List models request and click the “Send” button to send the request to the OpenAI API. The response from the API will be displayed in the “Response” section of Postman. The response status code should be 200 OK.

Creating well-researched and to-the-point content requires a lot of time and energy. If this was helpful and you wish to support me, please leave a comment, share, and press that 👏 a few times (up to 50 times). And consider subscribing to Medium.

You can now use Postman to make requests to the OpenAI API and use its various models and functions.

Remember to save any new requests in the OpenAI collection so that they can inherit the authentication information that is configured in the collection.

Using the Completions API from OpenAI

You can use the Completions API to get answers that are somewhat similar to the ChatGPT using the text-davinci-003 model. Let me show you how.

To use the Completion API from OpenAI in Postman, you will need to have a valid API key. If you haven’t done this yet, follow the steps above. Once you have an API key, you can use it to make requests to the Completion API using Postman. To do this, follow these steps:

Step 1. In Postman, click the “New” button to create a new request.

Step 2. In the “New Request” window, select “POST” as the HTTP method and enter the following URL: https://api.openai.com/v1/completions

Step 3. Save the request as “Create completion” inside the OpenAI collection. To do this, click on the “Save” button, select the OpenAI collection and rename the request. Finally, click on Save. This step is essential as the API key is stored only in the OpenAI collection.

Step 4. In the “Body” tab, select the “raw” option and enter the following JSON payload in the text area:

{
  "model": "text-davinci-003",
  "prompt": "What gifts should I buy for Christmas?",
  "max_tokens": 64,
  "temperature": 0.5
}

Step 5. Click the “Send” button to send the request to the OpenAI API. The response from the API will be displayed in the “Response” section of Postman, and the body should contain gift ideas.

If you want to learn how to use Postman for API testing, I have created an online course that starts with the basics of Postman and helps you understand the fundamentals of API testing. Learn more about the course.

A quick overview of OpenAI-specific terms and concepts

Now that the request works let’s clarify a few technical aspects. If you are new to AI, you are probably wondering about the parameters we have submitted in the request. Let’s briefly explain what this means.

Model

The text-davinci-003 model is a language generation model developed by OpenAI. It is capable of generating human-like text in a variety of styles and formats, including news articles, stories, and poems.

The text-davinci-003 model is based on the GPT-3 (Generative Pre-training Transformer 3) architecture, which uses transformer layers to process and generate text. It is trained on a massive dataset of internet text, including books, articles, and websites, and has been fine-tuned on specific tasks such as translation, question answering, and language modeling.

The text-davinci-003 model is one of the most advanced and powerful language generation models currently available and is capable of generating coherent and expressive text. It can be used for a variety of applications, including content creation, language translation, and conversational AI.

Prompt

A prompt is a piece of text or a set of instructions that is used as input to a language generation model. The model uses the prompt as a starting point to generate text based on the content and style of the prompt.

When making a request to the API, you can specify a prompt in the request payload, along with other parameters, such as the length and style of the generated text. The API will then use the prompt to generate text and return the result in the API response.

Token

In the context of natural language processing and language generation, a token is a unit of text that is used by a model to represent language input and output. Tokens can be words, punctuation, or other types of symbols, depending on the model and the task at hand.

Tokens are often used in language generation models to process and generate text. For example, a model might take a sequence of tokens as input and generate a new sequence of tokens as output. The model can use the tokens to represent the meaning and structure of the input and output text and to manipulate it in various ways.

I know this sounds a bit confusing, so allow me to give a practical example from the OpenAI documentation. A token is about 4 characters in English. If we scale it up a bit, 100 tokens are 75 words. This information is relevant to the cost of using the API. [2]

If you look closely at the response body in the previous example, you will see how many tokens have been used in the process.

So the property max_tokens is a way to control the output size and the associated cost. Each model also has a limit of tokens it can output, typically between 2048 and 4096 tokens.

Temperature

The temperature setting has nothing to do with the temperature of the CPU or GPU running the AI or the weather outside. It is actually an abbreviation for temperature sampling. As Ben Mann explains, “Temperature sampling is inspired by statistical thermodynamics, where high temperature means low energy states are more likely encountered.” [3]

Temperature is a value that is typically set between 0 and 1, with higher values indicating more random and creative output and lower values indicating more predictable and conservative output.

For example, a model with a high temperature might generate more diverse and unpredictable text, while a model with a low temperature might generate more repetitive and predictable text.

Is OpenAI free?

OpenAI seems like a free service, but it is not. OpenAI offers a fair, pay-as-you-go pricing model. You essentially need to buy tokens. For the Davinci model, 1,000 tokens cost around US $0.02.

However, when you create a new account, you get $18 in credits you can start using any OpenAI products in the first three months. [4]

Explore more endpoints

In this tutorial, we have just covered the basics of using the OpenAI API. If you look over the OpenAI API reference, you will see that there is so much more to be done. Let me know by posting a response below which topics or endpoints I should cover next.

Conclusion

I hope this tutorial helped you get started with the OpenAI API in Postman. Leave a comment in the section below if you have any questions. I would love to hear from you!

Thank you for sticking with this article until the end. If you enjoyed it, please leave a comment, share, and press that 👏 a few times (up to 50 times). It will help others discover this information, and maybe it will help someone else as well.

Follow me on Medium and YouTube if you’re interested in more tutorials like this.

References

OpenAI
ChatGPT
Postman
Openai Chatgpt
Openai Completions
Recommended from ReadMedium
avatarAlexander Nguyen
The resume that got a software engineer a $300,000 job at Google.

1-page. Well-formatted.

4 min read
avatarAbhay Parashar
17 Mindblowing Python Automation Scripts I Use Everyday

Scripts That Increased My Productivity and Performance

14 min read
avatarAndrew Zuo
Async Await Is The Worst Thing To Happen To Programming

I recently saw this meme about async and await.

5 min read
avatarChristopher Tao
Do Not Use LLM or Generative AI For These Use Cases

Choose correct AI techniques for the right use case families

7 min read
avatarVishal Rajput
Prompt Engineering Is Dead: DSPy Is New Paradigm For Prompting

DSPy Paradigm: Let’s program — not prompt — LLMs

11 min read
avatarDerek Johnson
I’m Unemployed for Over Two Years (as a software engineer)

In 2022, I worked on a contract as a software engineer at Apple. Apple dissolved our entire team right before the 2022 tech recession…

3 min read