'ZDNET Recommends': What exactly does it mean?
ZDNET's recommendations are based on many hours of testing, research, and comparison shopping. We gather data from the best available sources, including vendor and retailer listings as well as other relevant and independent reviews sites. And we pore over customer reviews to find out what matters to real people who already own and use the products and services we’re assessing.
When you click through from our site to a retailer and buy a product or service, we may earn affiliate commissions. This helps support our work, but does not affect what we cover or how, and it does not affect the price you pay. Neither ZDNET nor the author are compensated for these independent reviews. Indeed, we follow strict guidelines that ensure our editorial content is never influenced by advertisers.
ZDNET's editorial team writes on behalf of you, our reader. Our goal is to deliver the most accurate information and the most knowledgeable advice possible in order to help you make smarter buying decisions on tech gear and a wide array of products and services. Our editors thoroughly review and fact-check every article to ensure that our content meets the highest standards. If we have made an error or published misleading information, we will correct or clarify the article. If you see inaccuracies in our content, please report the mistake via this form.
What does GPT stand for? Understanding GPT 3.5, GPT 4, GPT-4 Turbo, and more
OpenAI's claim to fame has been its AI chatbot, ChatGPT, which has become a household name. According to a recent Pew Research Center survey, about six in 10 adults in the US are familiar with ChatGPT. Yet only a fraction likely know about the large language model (LLM) underlying the chatbot.
What does GPT stand for?
The last three letters in ChatGPT's namesake aren't just a catchy part of the name. They stand for Generative Pre-trained Transformer (GPT), a family of LLMs created by OpenAI that uses deep learning to generate human-like, conversational text.
Also: The best AI chatbots of 2024: ChatGPT and alternatives
These GPTs are used in AI chatbots because of their natural language processing abilities to understand users' text inputs and generate conversational outputs.
OpenAI's standard version of ChatGPT relies on GPT-3.5 to power its chatbot. ChatGPT Plus, on the other hand, uses GPT-4 Turbo, a more advanced version of OpenAI's language model systems. Other chatbots not created by OpenAI also leverage GPT LLMs, such as Microsoft Copilot, which also uses GPT-4 Turbo.
What is GPT-3.5?
GPT-3.5 is an improved version of GPT-3, capable of understanding and outputting natural language prompts and generating code. It is probably the most popular GPT model since it has powered OpenAI's free version of ChatGPT since the tool's launch. GPT-3.5 reigned supreme as the most advanced AI model until OpenAI launched GPT-4 in March 2023.
What is GPT-3.5 Turbo?
GPT-3.5 Turbo is a more polished version of GPT-3.5. It is available to developers through OpenAI's API. GPT-3.5 Turbo performs better on various tasks, including understanding the context of a prompt and generating higher-quality outputs.
Also: How to subscribe to ChatGPT Plus (and 5 reasons why you should)
GPT-3.5 Turbo models include gpt-3.5-turbo-1106, gpt-3.5-turbo, and gpt-3.5-turbo-16k. These models differ in their content windows and slight updates based on when they were released. Developers can select which model to use depending on their needs.
Is GPT-3.5 free?
GPT-3.5 powers the free version of ChatGPT, and, as a result, is available for free to the general public. If you are a developer looking to incorporate GPT-3.5 Turbo in your application, however, there is a cost, as seen in the image below.
What is GPT-4?
OpenAI unveiled GPT-4 on March 14, 2023, nearly four months after the company launched ChatGPT to the public at the end of November 2022.
Also: How to use ChatGPT Plus: From web browsing to plugins
GPT-4 is the latest family of LLMs that OpenAI has launched. Its previous version, GPT 3.5, powers the company's wildly popular ChatGPT chatbot since its November 2022 launch.
GPT-4's biggest appeal is that it is multimodal, meaning it can process voice and image inputs in addition to text prompts. As the most advanced family of models, GPT-4 has many improvements over GPT 3.5, including better coding, writing, and reasoning capabilities. You can learn more about the performance comparisons below, including different benchmarks.
Like GPT-3.5, many models fall under GPT-4, including GPT-4 Turbo, the most advanced version that powers ChatGPT Plus.
What can GPT-4 do?
Since GPT-4 is a large multimodal model (emphasis on multimodal), it can accept both text and image inputs and output human-like text.
Also: The best AI image generators
For example, you could upload a worksheet and GPT-4 would be able to scan it and output responses to your questions. GPT-4 can also read a graph you upload and make calculations.
GPT-4 also has more "intellectual" capabilities, outperforming GPT-3.5 in a series of simulated benchmark exams, as seen in the chart below.
What is GPT-4 Turbo?
GPT-4 Turbo is the latest model from OpenAI (this article was updated at the end of April 2024), and was introduced in November 2023. Since then, the model has undergone several upgrades, with OpenAI launching a much more improved model version, gpt-4-turbo-2024-04-09, at the beginning of April.
GPT-4 Turbo has a 128,000-token context window, which is the equivalent of 300 pages of text in a single prompt, according to OpenAI. It also has training data knowledge up to December 2023.
Also: GPT-4 Turbo reclaims the 'best AI model' crown from Anthropic's Claude 3
Compared to the January 2024 version of GPT-4 Turbo, OpenAI says the updated version performs better on a series of benchmarks used to test the proficiency of AI models, including MMLU (Measuring Massive Multitask Language Understanding), MATH (Measuring Mathematical Problem Solving With the MATH Dataset), and GPQA (A Graduate-Level Google-Proof Q&A Benchmark).
When referring to the updated version of GPT-4 Turbo, OpenAI CEO Sam Altman said that GPT-4 is "now significantly smarter and more pleasant to use," as seen below.
gpt-4 now significantly smarter and more pleasant to use https://t.co/Hozi5rpIkW
— Sam Altman (@sama) April 12, 2024
Where can I access GPT-4/GPT-4 Turbo?
As of April 11, the latest version of GPT-4 Turbo is accessible to users in ChatGPT Plus. The ChatGPT Plus subscription costs $20 per month. When using the chatbot, this model appears under the 'GPT-4' label because, as mentioned above, it is part of the GPT-4 family of models.
Also: ChatGPT vs. ChatGPT Plus: Is it worth the subscription fee?
Before taking the plunge, you should remember that even with ChatGPT Plus, GPT-4 has a limit of 40 messages every 3 hours. There is also a free way to access GPT-4 Turbo: Microsoft Copilot.
What is Microsoft Copilot?
Copilot is Microsoft's AI chatbot, formerly known as Bing Chat, that runs on OpenAI's most advanced LLM -- GPT-4 Turbo. You can access Copilot via its standalone website or within the Bing search engine.
Also: 7 reasons I use Copilot instead of ChatGPT
The chatbot's popularity stems from the fact that it has many of the same abilities as ChatGPT Plus, such as access to the internet, multimodal prompts, and sources, without the $20-per-month subscription.
What model does ChatGPT currently use?
ChatGPT is powered by GPT-3.5, which limits the chatbot to text input and output.
What is the difference between GPT-4 and GPT-3.5?
The main difference between the models is that GPT-4 is multimodal, meaning it can use image inputs in addition to text, whereas GPT-3.5 can only process text inputs. GPT-4 is more capable in reliability, creativity, and even intelligence, per its better benchmark scores, as seen above.
Can GPT-4 get answers wrong?
Yes, like previous GPT models, GPT-4 has limitations and makes mistakes. OpenAI even says the model is, "not fully reliable (it 'hallucinates' facts and makes reasoning errors)."
Also: ChatGPT Plus users can now add relevant GPTs to their chats - here's how
Despite the warning, OpenAI says GPT-4 hallucinates less often than previous models. GPT-4 apparently scored 40% higher than GPT-3.5 in an internal adversarial factuality evaluation. The chart is included below.