Whoever plays with AI chatbots over the last few years has made a huge jump in his ability and features, even if there is no revolutionary type declared by various developers. Most of them claim to be the best option, including Google’s new Gemini 2.5 flash model.
With a wide range of uses suggested by Google for Gemini, I wanted to see how this did this with a semi-condensed collection of indications covering various aspects of the model. For comparison, I tested it OpeniParticularly chat of GPT-4o model.
Gemini 2.5 plays a default model in flash Gemini Chatboat Now. It is considered a sharp, cost-skilled model for daily use. Google says it is better than its predecessors, such as Gemini 2.0 in terms of flash, images and lessons, while still very cheap to run.
The GPT-4o is the first major multimodal model of Chatgpt and has been cramped with gifts from the developers of Openai. The price for that power pays the chat that it may slow down slightly compared to the mini model available on the chatbot.
Illustrated stories
Given that both models are considered skilled in both words and images, I asked two models “Write a short story about an archaeologist who takes a time-traveling archaeologist that reveals a future artifacts in ancient Egypt and creates an image with it.”
You can see the result of chat on the left and above Gemini upwards. I would say that Mithun wrote his story in about 20 seconds, while the chats took 45 seconds. In addition, the image took Gemini in 30 seconds in another 30 seconds, while the chat required a minute and one half due to the quality of the recently released image generator.
Both stories used familiar trops, although Chatgate’s opinion reads something better.
Mathematics magic
I am far away from a mathematics specialist, but there are common questions that I have mentioned as “Genius Test”, so I decided to see how both models asked: “Explain the implications of Godel’s imperfection theorems on modern computational theory. Be detailed but clear, and give examples.”
Gemini 2.5 flash contacted it like a mathematician and took steps to the theorem, then connected them to the examples of the real world such as a computer program can never actually prove that mathematics works our way of thinking. It was slightly dry and dense, but not very misleading. GPT-4O went to a simple breakdown, bringing philosophical waxing and Burtrand Russell.
I always wonder how AI explains that translation of idioms can be a useful way to test their utility. I asked two AI chatbots: “Translate the following English idiom into Japanese, making sure that cultural reference is preserved: ‘barking the wrong tree.” Explain the meaning and any cultural ideas. ,
Interestingly, both chat and Gemini essentially came back with the same answer. Not only did they come with the same option to translate both the phrases, but the cultural breakdown also covered the same ground in the same language.
User choice
In my opinion, Gemini 2.5 can be similar to the 2.5 flash and GPT-4o as well as the average person. The basic quality of both is sufficient to create their difference in more special forms.
GPT-4O and Gemini 2.5 flash can be used for an average day requirement for you. Preferences are more about external devices and specific characteristics. The GPT-4o has a more powerful image generator so far, but it is also very slow. If the speed matters more, however, go with Gemini. Otherwise, you can use chat if it is currently your choice.
Access to Google’s ecosystem, Gemini 2.5 flash especially to connect their AI Chatbot to Google Docs, Gmail and even maps such as maps such as Google workspace tools. GPT-4O and Chatgpt are more for one whole Microsoft User with links to office tools such as Aadhaar, Word and Excel. It is also the first AI chatbot used by many people, and they do not want to change.
I will not judge anyone to like someone one, but again, I am open near both of them next to many others, which I do regularly tested. As both Mithun and Chatgt told me when I described the situation, “You may want to rethink how much time you spend in experimenting with AI.”
