We asked ChatGPT, Gemini, and Claude what they think of each other

The internet is full of promises about artificial intelligence tools that will change your life, but how do you know which one is really worth choosing for your daily tasks? Instead of relying on marketing promises, a new and intriguing experiment presented each of the four leading chatbots — ChatGPT, Gemini, Claude, and Perplexity — with a simple question: what is their true and unfiltered opinion of their competitors? The honest answers reveal exactly when each tool will save you, and when it might disappoint.

Strong in writing and information retrieval

Claude by Anthropic was praised by its competitors as the most talented and natural writer, excelling in analyzing long documents and deep thinking. However, competitors noted that it tends to be too cautious or verbose, and sometimes avoids giving direct answers. On the other hand, Perplexity was identified as an excellent research engine based on real-time web search that provides reliable sources. Its notable drawbacks are a lack of creativity, weak conversation memory, and a tendency for hallucinations in citations when trying to use it as a regular chatbot.

Google's Gemini offers powerful integration with the search engine and Google tools, alongside impressive multimodal capabilities. Despite this, its rivals claim that it suffers from a lack of personality, and that its answers can be too superficial, prioritizing the breadth of information over its depth. Its integration with Google services makes it particularly efficient for daily work, but less suitable for tasks requiring deep and complex thinking.

The general assistant and the bottom line

OpenAI's ChatGPT was crowned by its competitors as a "Swiss Army knife" — the best versatile tool for creation, coding, and structured tasks, benefiting from a rich ecosystem and advanced features. On the other hand, its weaknesses lie in a tendency to agree too quickly with the user instead of challenging them, and in attempts to "smooth over" knowledge gaps instead of admitting it does not know. Furthermore, it is less excellent in long-term collaborative projects compared to Claude, and its quality sometimes varies between different updates.

Bottom line, instead of looking for the "best" tool, you should match the chatbot to your task: