The AI names people encountered in 2024 were not all the same kind of thing. Some were applications, some were families of models, and others were platforms for building AI systems. Understanding that distinction is more useful than declaring one universal winner. The examples here explain the 2024 landscape, not a current ranking of subscriptions or models.
A chatbot is not one fixed model
ChatGPT is a product; GPT-4o is an example of a model associated with that period. OpenAI’s model documentation specifies text and image input for GPT-4o. Our previous claim that ChatGPT lacked multimodal capabilities was too broad and incorrect. Capabilities need to be checked for the particular model and product interface, rather than inferred from the brand name.
Multimodal means more than a text box
Google’s May 2024 Gemini documentation described Gemini 1.5 Pro and Flash accepting combinations of text, images, audio and video as inputs. That describes input capabilities, not a guarantee that every consumer feature, account or output format was identical. A model understanding an image is also different from a tool generating a new image.
Image-generation products such as Midjourney and DALL·E belong in a different comparison from a text-only writing task. Before comparing examples, decide whether you need an editable visual, accurate text in an image, a consistent subject or simply an exploratory concept.
An enterprise platform is another category
IBM describes watsonx.ai as tools and resources for building generative AI and machine-learning solutions, including access to different foundation models. Calling “IBM Watson” one analytics model obscures those choices. A business evaluating a platform must also consider its data, deployment and governance requirements.
Compare a task you can judge
Use a small set of representative, non-sensitive examples. Write down what a correct answer needs, check errors and omissions, and include the effort required to review the result. Do not judge a coding assistant only by its prose or an image tool by an unrelated reasoning puzzle. Recheck current availability and terms before paying: a 2024 capability list is historical context.
Sources
developers.openai.com — gpt 4o
blog.google — gemini gemma developer updates may 2024
