Materials and Methods: Both English and Turkish versions of 41 multiple choice questions related to ocular external diseases and cornea were applied to ChatGPT-3.5, Copilot, and Gemini artificial intelligence chatbots. These questions were translated into Turkish by a certified native speaker. The answers given by the chatbots were compared with the answer key and grouped as correct and incorrect. The success rates of chatbots were compared statistically.

Results: In the English version of the questions, ChatGPT-3.5 provided correct answers at 53.7%, Copilot at 43.9%, and Gemini at 51.2% accuracy (p=0.655). In the Turkish version of the questions, ChatGPT-3.5 provided correct answers at 48.8%, Copilot at 41.5%, and Gemini at 43.9% accuracy (p=0.794). There was no statistically significant difference between chatbots in answering the Turkish versions of the questions, although there were fewer correct answers in all three applications (p>0.05).

Conclusion: Although artificial intelligence chatbots are a promising tool for obtaining information, they need to be developed and their performance improved both in terms of their knowledge level and ability to interpret and translate the meaning in different languages. Keywords : ChatGPT-3.5, Copilot, English and Turkish, Gemini, Ocular surface diseases and Cornea