ChatGPT Plugins: Uses and Alternative Options

Key takeaways

ChatGPT plugins extended the capabilities of ChatGPT by connecting with external tools and services. Here are some important facts to know:

• OpenAI discontinued ChatGPT plugins in April 2024 [1].

• OpenAI’s custom GPT feature replaced plugins, offering similar functions. 

• You can develop a custom GPT to perform use case-focused tasks without any coding knowledge, but plugins offer more complex functionality. 

Explore what ChatGPT plugins were, the customizable GPT feature that replaced them, and how plugins compare to GPTs. If you’re ready to start learning more, enroll in the IBM Generative AI Fundamentals Specialization. You’ll have the opportunity to gain experience with large language models, prompt writing, and generative AI in as little as four weeks.

Understanding ChatGPT plugins

ChatGPT plugins were third-party tools that you could add to ChatGPT to expand its functionalities. Similar to browser extensions, you could add a plugin to ChatGPT to extend and customize its function without modifying its code. Plugins could help ChatGPT access current information, automate certain functions, and run complex computations. 

Third-party developers specializing in a service typically developed these plugins for ChatGPT to allow the model to communicate with their external platforms and perform specialized tasks. For example, a ChatGPT plugin for Zapier allowed the AI engine to interact with several apps, like Google Sheets and Salesforce. These plugins leveraged application programming interfaces (APIs) to allow ChatGPT to interact with third-party tools, databases, and other services, facilitating the integration of industry-specific features and customized functionalities for specific use cases.

By extending the capabilities of ChatGPT’s base models, plugins were intended to improve user experience and efficiency. However, ChatGPT plugins were only available for GPT-4 models that had access to the web, which meant lower GPT models couldn’t integrate plugins. Additionally, since third-party vendors offered most of these plugins, there remained the risk of sensitive information getting exposed. 

Are ChatGPT plugins no longer available?

Yes, ChatGPT plugins are no longer available. OpenAI discontinued plugins in April 2024, preventing users from continuing chats that had integrated plugins [1].

What replaced ChatGPT plugins?

OpenAI introduced custom GPTs that can perform similar functions as plugins. You can customize your GPTs by adding instructions and documents or even integrating third-party services. 

Custom GPTs utilize RESTful APIs to connect to external services. Using natural language processing, the GPT understands the user’s question, determines which API would be most relevant, executes the API through function calls, and provides a response in natural language as well. This allows the GPT to retrieve information from or perform an action in an external application by converting text into a JSON schema for the API call. What's more, you can simply specify the schema, add an authentication mechanism, and provide instructions in natural language to the GPT, and the model will do the rest itself. 

ChatGPT’s custom GPT features enable anyone to create a GPT model for their specific use without requiring any coding. By simply chatting with ChatGPT, adding instructions in the form of prompts that guide your custom GPT’s behavior, providing documents relevant to your use case, and specifying the actions you want it to perform, you can create a customized GPT tailored to your needs.

This feature makes GPTs a powerful alternative to plugins, as you no longer have to rely on the plugins available from third-party vendors for specific tasks. Instead, you can create your own custom GPT that can call on these functions from external sites and perform specialized tasks. In fact, you might not even need to connect to external APIs for many of the functions that were previously handled by plugins, as the newer GPT models already come with advanced capabilities in contextual awareness and response accuracy.

Even if you don’t want to build your own GPT, you can browse the GPT Store to find millions of GPTs contributed by users and OpenAI’s partners that perform a diverse range of functions. This might have also contributed to the discontinuation of plugins, as several companies that previously offered ChatGPT plugins now have GPTs that perform the same functions.

Are GPTs safe to use?

While GPTs are generally safe to use, some challenges do exist in terms of security and accuracy. GPT models can hallucinate and present incorrect information confidently. Although newer GPT models are more advanced, the possibility of hallucination still exists, which can not only contribute to misinformation spreading but also lead to inaccurate outcomes for predictions, which is especially relevant in areas like health care and law. 

In terms of security, custom GPTs may have underlying vulnerabilities that potential attackers could target. A 2025 study found that 95 percent of the 14,904 custom GPTs examined lacked sufficient security protection [2]. The ability of custom GPTs to interact with external platforms can lead to potential security risks, like accidental leakage of sensitive information or credentials, and third-party APIs accessing and collecting user data. OpenAI does not control how these third-party services use your data; therefore, it recommends that you only use trustworthy external APIs.

Pros and cons of GPTs compared to plugins

A significant benefit of custom GPTs over plugins is that anyone can create them without coding knowledge by having a conversation with ChatGPT. In contrast, plugins were third-party APIs created by an external vendor outside of ChatGPT. This means that while you didn’t need to be a developer to create your GPT, you did need coding knowledge to create a plugin. However, this also means that plugins sometimes offered more complex functionality than GPTs, which are more suited for specialized tasks.

Some other differences in their benefits include:

• You can only use one GPT chatbot at a time, while you could integrate up to three plugins for a single chat. 

• GPTs are more cost-effective than plugins since they can perform a wide variety of simple tasks without the need for complex development.

• You can customize your instructions and conversational styles for your GPT to tailor its behavior for specific use cases; however, you can’t modify the instructions for plugins.

• Developing and deploying a custom GPT is generally a much quicker process compared to plugins, which require more maintenance.

Is there any AI better than ChatGPT?

While you can’t definitely say a single AI is better than another, since the best option for you would depend on your needs, you could consider the following ChatGPT alternatives based on the function you require:

• Claude AI: To generate human-like text and converse in a smooth, easy-to-understand manner

• Microsoft Copilot: To generate creative images and integrate with other Microsoft apps

• Google Gemini: To analyze and generate text, code, and images and connect with other Google apps 

• Meta’s Llama 4: A free, open-source tool for advanced content creation on Meta’s social media apps

• Perplexity AI: A solid alternative for online research, real-time web browsing, and producing cited responses

Careers and skills connected to GPTs

You typically won’t need any specific technical skills to make a GPT. However, you’ll need knowledge about the industry or task you’re making your GPT for, as it’ll help you understand what type of data you’ll need to use to build your model. Another skill that can benefit you in creating custom GPTs is prompt engineering, which will help you craft detailed prompts to generate more effective responses from your GPT. To better understand how to integrate external services with your GPT, it may also be helpful to have some knowledge of APIs. Lastly, a fundamental understanding of large language models (LLMs), artificial intelligence (AI), and natural language processing can help you better communicate with ChatGPT.

You may need to build or work with custom GPTs in a range of careers, not only those related to AI, such as marketing product management, or financial analysis. However, some AI-related roles that might require you to work more closely with GPTs, as well as their median US salaries, include:

• Prompt engineer: $125,000 [3]

• AI trainer: $84,000 [4]

• LLM engineer: $153,000 [5]

• AI product manager: $188,000 [6]

• AI consultant: $203,000 [7]

All salary information represents the median total pay from Glassdoor as of October 2025. These figures include base salary and additional pay, which may represent profit-sharing, commissions, bonuses, or other compensation.

Read more: How to Become a Prompt Engineer: Duties, Skills, and Steps

Explore AI learning opportunities

Want to keep the momentum going? Learn more about the latest trends and in-demand skills in the field of AI by subscribing to our LinkedIn newsletter, Career Chat! Or if you want to keep exploring AI careers and concepts, check out the following free resources:

• Take the quiz: AI Career Quiz: Is It Right for You?

• Watch on YouTube: Stop Using AI Like Google! Expert's #1 Prompting Tip

• Bookmark for later: Artificial Intelligence Glossary: Learn AI Vocabulary

Whether you want to develop a new skill, get comfortable with an in-demand technology, or advance your abilities, keep growing with a Coursera Plus subscription. You’ll get access to over 10,000 flexible courses. 

Build job-ready skills with Coursera Plus

Start 7-day free trial

• 
• 
• 
• 

Start 7-day free trial