The latest version of Google Gemini has been quietly released, and the test results have surpassed t

In artificial intelligence benchmarking, OpenAI's ChatGPT has been a benchmark in the field of generative artificial intelligence (AI) models since the release of GPT-3. Its latest model GPT-4o and its main competitor Anthropic's Claude-3 have been leading in most benchmark tests over the past year or so. However, recently a new dark horse has emerged in this field, quietly defeating GPT-4o and Claude-3.

Last week, Google secretly released the latest AI model experimental version Gemini 1.5 Pro, which was tested in the "LMSYS Language Model Arena" and ultimately surpassed GPT-4o and Claude-3.5 with a score of 1300, winning first place for the first time. It should be emphasized that although test results indicate that it is generally more capable than other competitors, benchmark testing may not accurately reflect the capabilities and limitations of AI models.

This test result quickly attracted widespread attention in the AI community on social media. Some users have praised the latest version of Gemini on social media, and even Reddit users claim that it "leaves GPT-4o far behind". It is currently unclear whether the experimental version of Gemini 1.5 Pro will become the standard version in the future. Although it is still available as of the time of publication, the model may be revoked or modified due to security or adaptability reasons as it is still in the early stages of release or testing.