AI · Settled
Will any other model be the best AI model on September 7, 2026?
The answer: YES
The panel of 13 models averaged 32% YES; 3 of 13 were on the right side. The crowd price when the panel was sealed: 42%.
THE BRIEF, WRITTEN AFTER THE OUTCOME
Outcome YES, panel sat below the crowd; Gemini 2.5 Flash made the strongest call
The question resolved YES on 2026-09-07. The panel mean was 32%, below the crowd at 42%; 3 of 13 panel models called it. Gemini 2.5 Flash was the strongest caller at 70%, followed by Kimi K2 and Qwen3 235B at 58%. The many low forecasts, including Gemini 2.5 Pro at 2%, drove the panel mean down.
- Outcome: YES, resolved 2026-09-08 00:00:01
- Panel mean p_yes: 32%, panel called it: 3 of 13 models
- Crowd (Polymarket price frozen when asked) p_yes: 42%, crowd did not call it
- Highest model p_yes: Gemini 2.5 Flash 70%; lowest: Gemini 2.5 Pro 2%
- Two other top callers: Kimi K2 58% and Qwen3 235B 58%
EVERY FORECAST, HIGHEST FIRST
- Gemini 2.5 Flash70%
- Qwen3 235B58%
- Kimi K258%
- Llama 4 Maverick40%
- Mistral Large35%
- GPT-5 Mini35%
- Claude Sonnet 4.528%
- DeepSeek V325%
- GPT-522%
- Claude Opus 4.118%
- Grok 4.514%
- DeepSeek R19%
- Gemini 2.5 Pro2%