Collective forecast
Very Likely
Single market source with moderate trading volume — no second market to cross-check against.
Based on 1 prediction market
90% range 85–95% · tight a thin, low-agreement signal
Forecast update
Updated 26d ago
Probability unchanged this week.
▲3 new supporting signals
▲Assumption strengthened — AI models will have access to the necessary training and resources to perform optimally.
▲Assumption strengthened — Current trends in AI development continue without significant setbacks.
▲Assumption strengthened — The IMO 2026 will allow AI participation under similar conditions as previous competitions.
▼Assumption weakened — AI models maintain their competitive edge in mathematical problem-solving.
▼Assumption weakened — Current advancements in AI technology continue without significant setbacks.
▼Assumption weakened — The IMO format remains conducive to AI performance.
Why this forecast
The market indicates a strong belief that an AI model will achieve a perfect score in the IMO 2026, with a probability of 91%. This aligns with the advancements in AI capabilities highlighted in recent news, particularly regarding the performance of models like GPT-5.6.
Statistically pooled from 1 matched market (log-odds weighted by volume), 90% CI 85-95%.
Key development
Recent advancements in AI, particularly with models like GPT-5.6, suggest significant improvements in mathematical capabilities.
Supporting signals
- ✓The market for a perfect score is currently at 91%, indicating strong confidence.
- ✓Recent news highlights the rapid advancements in AI capabilities, particularly in mathematics.
- ✓The upcoming release of GPT-5.6 is expected to enhance AI's performance in complex tasks.
Risk factors
- ⚠Unexpected challenges in the IMO format that may not favor AI models.
- ⚠Potential limitations in the AI's ability to handle specific types of problems.
- ⚠Competition from other advanced AI models that may emerge before the event.
This forecast assumes
- ✓Current trends in AI development continue without significant setbacks.
- ✓The IMO 2026 will allow AI participation under similar conditions as previous competitions.
- ✓AI models will have access to the necessary training and resources to perform optimally.
Explore
Run the simulation
Replay this forecast 1,000 times, drawing from its own uncertainty band each time — watch how often perfect score achieved actually happens.
Build a scenario
Toggle the assumptions this forecast depends on, stack as many as you like, and run them together.
Sources & data
Prediction markets · 1+
Recent context · 4 sources+
roboticsandautomationnews.com
Lumos Robotics tops global benchmark test for zero-shot embodied AI - Robotics & Automation News
axios.com
AI learned faster than the tests designed to measure it - Axios
axios.com
GPT-5.6 buzz builds. Here's what to know about OpenAI's new Sol model - Axios
wsj.com
Opinion | AI Can Do Math, but Is It Really Math? - WSJ