What will be true of the first model to cross 1400 on lmarena.ai?
16
1kṀ4694
resolved Mar 1
Resolved
NO
Gemini Exp
Resolved
NO
ChatGPT 4o
Resolved
NO
o1
Resolved
NO
Gemini 2.0
Resolved
NO
Claude 3.5 Opus
Resolved
NO
Claude 4
Resolved
YES
Grok
Resolved
NO
OpenAI model code named Orion
Resolved
NO
GPT 5

Will resolve if a model stays at or above 1400 for a week and has a 95% CI with a lower bound of at least 1395 at the end of that week (somewhat arbitrary criteria to ensure the score is based on a sufficient amount of votes)

Will N/A if they change the scoring significantly so that a current model passes 1400.

Current rankings (11/22/24):

  1. Gemini Exp 1121: 1365

  2. ChatGPT 4o Latest (2024-11-20): 1360

  3. Gemini Exp 1114: 1343

  4. o1 preview: 1334

  5. o1 mini: 1308

  6. Gemini 1.5 Pro-002: 1301

  7. Grok 2 0813: 1289

  8. Yi Lightning: 1287

  9. GPT 4o 2024-05-13: 1285

  10. Claude 3.5 Sonnet (20241022): 1282

  • Update 2025-24-01 (PST): - If a Deepseek model is first to cross 1400, all will resolve to NO (AI summary of creator comment)

AI
Technology
Technical AI Timelines
OpenAI
LLMs
Get
Ṁ1,000
to start trading!

🏅 Top traders

#NameTotal profit
1Ṁ128
2Ṁ128
3Ṁ66
4Ṁ66
5Ṁ14
Sort by:

Resolution criteria

"stays at or above 1400 for a week and has a 95% CI with a lower bound of at least 1395 at the end of that week"

early Grok 3 is over 1400 as of 02-16, so will need to maintain that rating until 02-22 to resolve to yes

bought Ṁ50 NO

If a Deepseek model is first to cross 1400, all of these will resolve NO

@ChinmayTheMathGuy also o3

People are also trading

Will any AI model achieve > 40% on Frontier Math before 2026?
35% chance
First model series to cross 1500 on lmarena.ai?
Top score on codeforces by an AI model at the end of 2025
Model at 1600 Elo on lmarena before 2026?
4% chance
Who will have the highest ranking model on web.lmarena.ai by EOY 2025?
Will any AI model score >80% on Epoch's Frontier Math Benchmark in 2025?
5% chance
Will an AI model outperform 95% of Manifold users on accuracy before 2026?
19% chance
Will an AI model achieve superhuman ELO on Codeforces by the 31 December 2025?
55% chance
Will a single model achieve superhuman performance on all OpenAI gym environments by 2025?
25% chance
Will the top model by OpenAI rank 3rd (or lower) behind 2 other model families at any point before 2026?
66% chance
© Manifold Markets, Inc.•Terms•Privacy