Home ยป Gemini 1.5 Pro’s Latest Trial Version Outperforms GPT-4o in Chatbot Arena Test Last Week

Gemini 1.5 Pro’s Latest Trial Version Outperforms GPT-4o in Chatbot Arena Test Last Week

LMSYS is a platform that ranks chatbots based on their ability to provide responses from multiple chatbots for users to choose the most suitable one. The latest results show that the test version 0801 of Gemini 1.5 Pro outperformed GPT-4o and claimed the top spot for the first time.

The 0801 version model can be used in AI Studio but has not yet been widely adopted. Meanwhile, Gemini Advanced ranks fourth along with Claude 3.5 Sonnet and Llama 3.1 405B, marking the first time an open-source model has achieved such a high ranking.

Although the overall ranking may be number one, individual topics may vary. For example, GPT-4o still prevails when faced with difficult questions, while Claude 3.5 Sonnet is the top choice for programming.

Google previously held the highest rank on LMSYS at the beginning of the year, placing second when using Gemini Pro.

Source: LMSYS

TLDR: Gemini 1.5 Pro test version 0801 surpasses GPT-4o in LMSYS chatbot rankings, marking the first time an open-source model has achieved such a high ranking.

More Reading

Post navigation

Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

OpenAI Launches ChatGPT Edu for Educational Institutions: A feature-rich Enterprise-Level Tool at a Budget-Friendly Price

GPT-4o’s Thai Language Tokenizer Test Yields Remarkable Efficiency

The AI Battle of 2024: OpenAI Unveils GPT-4o to Challenge Google’s Dominance in a Single Day