TOTAL VOLUME:

$116b

24H VOL:

$91,034,583

24H TRANSACTIONS:

1,351,258,445

OPEN INTEREST:

$1,137,445,785

321,343

Markets across

31,490

events

MATCHED EVENTS:

3,478

PLATFORM COVERAGE:

5

Polymarket:

42%

VS.

Kalshi:

58%

BETA

Science and Technology

AI model scores ≥ 90% on FrontierMath Benchmark before 2027?

Science and Technology

polymarket

kalshi

AI model scores ≥ 90% on FrontierMath Benchmark before 2027?

Total volume:

$168,301

Volume 24h:

$56

180%

Liquidity:

$2,523

4%

Open interest:

$26,695

0%

AI model scores ≥ 90% on FrontierMath Benchmark before 2027?

- Polymarket

AI model scores ≥ 90% on FrontierMath Benchmark before 2027? - Polymarket

81%

-10%

1W

News

Positive

Negative

Neutral

Hover marker for details

Aug 18

Aug 19

Aug 20

Aug 21

Aug 23

Aug 24

Aug 25

Vol.

$1.3k

·

Resolves Dec 31, 2026

AI model scores ≥ 90% on FrontierMath Benchmark before 2027?

Amount

$

Trade on

polymarket

Trade on Polymarket

At 81¢ buys you 123 shares | Odds: 80% Total Payout: $123 | Net Profit: $23 Multiplier: 1.23x | ROI: 23% | APY: 82% 127 days to resolution

You will be redirected to the platform to complete this trade.

Markets

Outcome

Trade

Chance %

Price

Spread

Liquidity

Volume

24h

7d

Open Interest

Ends in

Result

title

trade

chance

price

spread

liquidity

volume

volume24pers

volume7pers

openInterest

endDate

unifiedStatus

polymarket

AI model scores ≥ 90% on FrontierMath Benchmark before 2027?

80%

Yes 81¢

No 21¢

2¢

$2,523

$118,627

0.05%

1%

N/A

4mo 1w

active

kalshi

Before 2027

97%

Yes 100¢

No 100¢

100¢

N/A

$7,970

N/A

N/A

$4,032

4mo 1w 2d

No

kalshi

Before September

96%

Yes 100¢

No 100¢

100¢

N/A

$16,039

N/A

N/A

$9,227

4mo 1w 2d

No

kalshi

Before October

96%

Yes 100¢

No 100¢

100¢

N/A

$9,337

N/A

N/A

$3,989

4mo 1w 2d

No

kalshi

Before December

95%

Yes 100¢

No 100¢

100¢

N/A

$8,790

N/A

N/A

$4,541

4mo 1w 2d

No

kalshi

Before November

91%

Yes 100¢

No 100¢

100¢

N/A

$7,538

N/A

N/A

$4,906

4mo 1w 2d

No

Total markets: 6

Description

This event group tracks whether any state-of-the-art AI model will achieve a score of 90% or higher on the FrontierMath Benchmark before the end of 2026. The outcome depends on the performance of AI models on complex mathematical problems.

PredictionHero - Resolution Divergence Alerts (RDA)

Divergence Detected

Issue: Timing mismatch between platforms regarding when and how the event is resolved.Hero tip: Focus on AI breakthroughs occurring before December 31, 2026, for Polymarket. For Kalshi, any solution before January 1, 2027, is sufficient.

Critical divergence points:

Polymarket: Resolves to Yes if a SOTA AI model achieves a score of 90% or greater on the FrontierMath Exam by December 31, 2026.
Kalshi: Resolves to Yes if any AI solves at least one Frontier Math: Open Problem before January 1, 2027.

Our PredictionHero Resolution Divergence Alerts (RDA) are there to help users identify potential differences across platforms. They do not replace or supersede the official rules and description of any prediction market. Users are solely responsible for reviewing and understanding the applicable rules and resolution criteria before placing any trade or bet. If you notice a potential inconsistency, discrepancy, or error in an alert, please report it to our team so we can review and improve the accuracy of our data.

Show more

Polymarket

This market will resolve to "Yes" if a state-of-the-art (SOTA) AI model achieves a score of 90% or greater on the FrontierMath Exam by December 31, 2026, 11:59 PM ET. Otherwise, the market will resolve to "No". The primary resolution source will be information from EpochAI however a consensus of credible reporting may also be used.

Kalshi

The event evaluates whether any AI system solves at least one designated 'Frontier Math: Open Problems' between issuance and the respective deadline for each market. All markets share identical outcome criteria but differ only in their resolution dates, ranging from September 2026 to January 2027. For all markets, two specific problems—a Ramsey-style problem on hypergraphs and The 2-adic Absolute Galois Group—had already been solved by AI at issuance; solving these same problems again with different AI models does not satisfy market conditions. Resolution depends solely on whether a distinct, new Frontier Math problem is solved by AI within the applicable timeframe for each individual market.

Frequently asked questions

The dashboard for the FrontierMath AI benchmark market aggregates data from multiple venues, showing how traders across platforms price the likelihood of an AI model achieving at least a 90 percent score on the FrontierMath Benchmark before the end date. It displays current probabilities, volume of $168,225, and 24-hour volume of $21,478, offering a consensus view of market sentiment toward this technological milestone.

On Polymarket, this market currently reflects a strong probability of success, while some analyst forecasts may vary based on different assumptions about computational advances and benchmark difficulty. The gap between market odds and expert opinions can hinge on recent breakthroughs, funding announcements, or shifts in the research community’s focus. Comparing these sources helps identify where crowd-sourced intuition diverges from established research timelines.

This market resolves around Dec 31, 2026, with the outcome confirmed once the event is verifiable from credible public reporting. If any AI model publicly demonstrates a score of 90 percent or higher on the FrontierMath Benchmark before that date, the affirmative side wins; otherwise, the negative side is selected. No additional criteria or private data are required for resolution.

Key signals include major research publications, announcements from leading AI labs about new training techniques, or public benchmark results shared on platforms like arXiv or GitHub. Significant investments in computational resources, collaborations between tech giants, or even regulatory changes affecting AI development could also shift probabilities. Any demonstration — even incomplete — of near-90 percent performance may cause rapid repricing across venues.