Will any AI consistently get a perfect score on Maxim Lott's offline IQ test by the end of 2027?

Ṁ1kṀ4.2k

2028

94%

chance

ALL

https://www.maximumtruth.org/p/deep-dive-ai-progress-continues-as

Vision mode. Must get a perfect score on 3 consecutive tests, as displayed by the website. (Open to tweaking if the website changes its display or testing procedure.)

Technology

OpenAI

AI Impacts

Technical AI Timelines

Get

1,000

to start trading!

People are also trading

Will AI pass the Longbets version of the Turing test by the end of 2029?

54% chance

Will any AI model score 60+ on the Artificial Analysis Intelligence Index by November 20, 2026?

58% chance

Will an AI be capable of achieving a perfect score on the Putnam exam before 2027?

89% chance

Will AIs beat human experts in question-answering on the GPQA benchmark before January 1st, 2027?

98% chance

Will any AI achieve a score of 25% on ARC-AGI-3 by the end of 2026?

99% chance

Will an AI be capable of achieving a perfect score on the Putnam exam before 2028?

92% chance

Will an AI be capable of achieving a perfect score on the Putnam exam before 2030?

96% chance

Will any AI model score above 95% on ARC-AGI-2 by end of 2026?

99% chance

Will an AI achieve a perfect score on the Miklós Schweitzer Competition before 2028?

83% chance

Will an AI achieve a perfect score on the Miklós Schweitzer Competition before 2035?

81% chance

Sort by:

bought Ṁ196 YES

@IsaacKing Resolves YES

https://trackingai.org/test.html?iq-test=offline&iq-vision=1&iq-ai=%5B%22Claude-5-Opus-Max-Vision%22%2C%22Offline+Test%22%2C%22Claude-5-Opus-Max-Vision%22%5D#iq-test-chart

Claude-5.5 Opus MAX (Vision)

PROVision

(claude-opus-5.5, MAX thinking, with vision/image read)

IQ Score135

Averaged Over:

October 18, 2026 (IQ: 136)

October 11, 2026 (IQ: 136)

October 4, 2026 (IQ: 136)

September 28, 2026 (IQ: 136)

September 21, 2026 (IQ: 130)

Avg. Time / Response38.7 s

Cost / Response$0.0079

Raw Score (% correct): 16 out of 16 (100%)

Test: Offline Test

By the way @MaximLott are you concerned about the offline test making it into training data from your conversations? Most companies are presumably using those as training data for their next models.

In your case this might not be that big a deal since you're not providing correct answers I suppose.