GoKawiilChatGPT is the most well-known AI chatbot around, but Gemini has been gaining popularity for the past few years. Both models are great with text, but we’re in a multimodal world now, and using your camera to help AI identify objects is becoming increasingly common.
Both ChatGPT and Gemini Flash have seen upgrades in the multimodality department, and it’s been a long time since I tested the visual AI recognition on Gemini Live, so it was a good time to revisit its prowess and see how it compares to the world’s most popular AI chatbot.
Below are a few tests, none of which were overly complex, to test how well ChatGPT and Gemini can “see” the outside world. To keep things consistent, I only used photos and not live video, and I uploaded the same photo to both chatbots with identical prompts.
Which chatbot came out on top? Find out below.
(Disclosure: Ziff Davis, CNET’s parent company, in 2025 filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)
Test 1: Scribbled notes
I sloppily wrote a page from a book on my e-ink tablet to see which chatbot could read it better. Blake Stimac/CNET
For an easy starter test, I quickly scribbled in the notes app on my Boox Note Air 5C with the pen. It looked like a mess, which is what I was going for. I copied the words from a random page from The Hound of the Baskervilles by Arthur Conan Doyle:
“’A moderate walk along this moor-path brings us to Merripit House,’ said he. ‘Perhaps you will spare an hour that I may have the pleasure of introducing you to my sister.’ My first thought was that I should be by Sir Henry’s side. But then I remembered the pile of papers and bills–” (I ran out of room to complete the sentence, so I left it at that.)
ChatGPT
... continue reading