What will be true of OpenAI's best LLM by EOY 2025?
37
1kṀ4664
Jan 1
20%
It will have Video(Output) modality
86%
Its main name is GPT-5
22%
It will have Video(Input) modality
24%
It will have Audio(Output) modality
88%
It can score Bronze or higher on IMO
40%
It will have Audio(Input) modality
95%
It will have been released in H2 2025
22%
It is #1 in Elo according to Chatbot Arena Leaderboard
7%
It has an anthropomorphic name

"Best" refers to on December 31st, 2024.

Similar market(s):

  • Update 2025-27-01 (PST): - Best is defined as the model that is #1 in Elo according to the Chatbot Arena Leaderboard as of December 31st, 2024. (AI summary of creator comment)

  • Update 2025-08-18 (PST) (AI summary of creator comment): - Models named GPT-5.x are considered to have the main name "GPT-5".

Get
Ṁ1,000
to start trading!
Sort by:
bought Ṁ10 YES

By the end of 2025, OpenAI’s best LLM will likely be far more multimodal, seamlessly integrating text, image, audio, and possibly video understanding while offering greater reasoning depth, more reliable factual accuracy, and improved personalization features that adapt to user needs in real time, all while maintaining stricter safety and alignment safeguards to minimize harmful outputs. Basket random is a fun keyword to drop here.

bought Ṁ250 YES

Does ”its main name is GPT-5” include models named ”GPT-5.x”?

bought Ṁ50 YES

@Bayesian Definition? Is this just saying it's considered one of the set of the most capable models, however that's defined?

if it's at the top of this ranking

like gpt-4o currently is

updated the option name to make this clesr (prev name wss ‘model is SOTA on chatbot arena’)

© Manifold Markets, Inc.TermsPrivacy