AI Face-Off

ChatGPT and Perplexity face off in our first round of Tom's Guide's annual AI Madness championship.

I asked ChatGPT, Claude and Gemini to compete in seven humor challenges — from roasting me to writing absurd jokes. The results reveal which AI is actually the funniest.

I asked Claude to write detailed prompts and used them to test ChatGPT against Nano Banana across seven image styles — from photorealism to fantasy. The results were surprisingly uneven.

 A simple one-word prompt can turn friendly chatbots into brutally honest critics — I tested it on ChatGPT and Claude to see which AI gives the toughest feedback.

Across three tests, I concluded which chatbot (ChatGPT and Perplexity) is better at handling my research prompts

I tested Gemini 3 Flash and Claude Sonnet 4.6 with 7 real-world prompts to see which AI assistant performs better for reasoning, planning, writing and creativity.

I took a break from ChatGPT to put Perplexity and Claude through a variety of tests to see which AI tool emerged as the better alternative

I tested ChatGPT-5.2 against Nano Banana 2 against each other to see which chatbot can generate the best-looking AI images based on my descriptive prompts

Can you trust AI for breaking news? We stress-tested Claude, ChatGPT and Gemini on the March 2026 Iran strikes. Which AI overfabricated —and which stayed honest.

ChatGPT and Claude's default models battle it out in challenges that test every day uses such as writing, reasoning and explanation to find the overall winner.