Today, OpenAI announced the release of an updated GPT-4o model with a few notable improvements. This latest GPT-4o model comes with improved creative writing ability that can deliver more natural, engaging, and tailored writing with better relevance and readability. OpenAI also claims that this new model is better at working with uploaded files, providing deeper insights and more thorough responses.
This new model is already live for ChatGPT users around the world. For developers, this new GPT-4o model is available under the following names:
- gpt-4o-2024-11-20 (API)
- chatgpt-4o-latest (API)
Both of the above models still come with a 128,000-token context window, 16,384 maximum output tokens, and training data up to October 2023.
With this new model, OpenAI also reclaimed the No. 1 rank in Chatbot Arena, an open platform for crowdsourced AI benchmarking. The ChatGPT-4o (20241120) was tested in Chatbot Arena as "anonymous-chatbot" over the past week, and it gathered around 8,000+ community votes.
As per the rankings, ChatGPT-4o (20241120) has surpassed the Gemini-Exp-1114 model, which was released just last week, with a score of 1361. According to the community scores, the latest GPT-4o delivered remarkable improvements in creative writing; its score increased from 1365 to 1402. Find the overall ranking changes below.
- Overall: #2 → #1
- Overall (StyleCtrl): #2 → #1
- Creative Writing: #2 → #1
- Coding: #2 → #1
- Math: #4 → #3
- Hard: #2 → #1
Latest ChatGPT-4o remains #1 with Style Control, and improvement across the board. pic.twitter.com/ihpGDeL9RG
— lmarena.ai (formerly lmsys.org) (@lmarena_ai) November 20, 2024
OpenAI"s updated GPT-4o model marks a shift from their previous o1 series, which focused on logical reasoning. GPT-4o prioritizes creative writing, demonstrating OpenAI"s commitment to exploring diverse capabilities in AI models. This is a refreshing change of pace from the industry"s current focus on math and coding.