
Which Millennium Prize Problem will AI solve next?
PM $672.4K · K 11.5K contracts
2
11

- Reported volume
- PM $672.4K · K 11.5K contracts
- Events
- 2
- Markets
- 11
Current odds
Snapshot from
As of Nov 22, 2026, 3:00 PM UTC, Riemann Hypothesis is the highest-priced outcome on both platforms, at 76% on Polymarket and 94.5% on Kalshi.
Polymarket
- Riemann Hypothesis76%
- Yang-Mills7.6%
- No solution by Dec 31, 20273.5%
- Birch and Swinnerton-Dyer3.3%
- Hodge Conjecture3%
Kalshi
- Riemann Hypothesis94.5%
- Yang–Mills Existence and Mass Gap Problem5.5%
- Hodge Conjecture3.5%
- Birch and Swinnerton-Dyer Conjecture2.5%
- P versus NP Problem2.5%
Yes prices. They are not order quotes; check each platform's rules, fees and live prices before trading.
AI Analysis
Trader mode: Market mechanics, assumptions and risks
PM $672.4K · K 11.5K contracts
Reported volume · provider units
11
Saved contracts
0
Same-question pairs
Source snapshot: Nov 22, 2026, 3:00 PM UTC
Snapshot observed Nov 22, 2026, 6:00 AM UTC. This analysis describes that snapshot and is not a live news update.
What the contract measures
This event asks which Millennium Prize Problem will be solved next, but the two platforms word it differently. Kalshi's rules state that "If X is the first Millennium Prize Problem solved after Issuance, then the market resolves to Yes," under the heading "Before 2045" [1]. A problem counts as solved at the earliest of three conditions: a complete proof, disproof, or counterexample is published in a peer-reviewed journal; the Clay Mathematics Institute or another recognized authoritative academic body recognizes the resolution in writing; or at least two Source Agencies report it [1].
Polymarket uses a different structure. Its market resolves according to "the first qualifying Millennium Prize Problem for which an eligible AI lab announces a solution by December 31, 2027, 11:59 PM ET" [2]. Only announcements by eligible AI labs qualify, and the Polymarket rules state that "No further confirmation from Clay Math Institute or any other organization is required" [2]. Navier-Stokes is excluded [2].
Reading the snapshot
The saved data was observed at Nov 22, 2026, 6:00 AM UTC and retrieved at Nov 22, 2026, 8:07 AM UTC. These figures describe that moment only and are not live prices.
Kalshi's five saved markets each cover one named problem, with a 2045-01-01 end date in the saved data [1]. Riemann is priced at 0.91 Yes, Yang-Mills at 0.055, Hodge at 0.035, and both P vs NP and Birch and Swinnerton-Dyer at 0.025. Polymarket's six saved markets cover a different question and window. Riemann is priced at 0.7805 Yes, Birch and Swinnerton-Dyer at 0.079, Yang-Mills at 0.076, Hodge at 0.041, and P vs NP at 0.0085. Its sixth market, "Will no Millennium Prize Problem be solved by AI by December 31, 2027?", is priced at 0.0285 Yes [2].
Prices are snapshots, not executable quotes or verified probabilities. Because the wording, windows, and eligible solvers differ, the Kalshi and Polymarket figures should not be averaged or treated as one estimate. Reported volumes are also not comparable: Kalshi reports 11,517 contracts and Polymarket reports $597,911, in different units. Neither figure indicates liquidity or the number of traders.
Both platform samples are complete for these saved events, with no partial sample flagged. That does not establish that the providers list every possible real-world outcome.
What remains uncertain
The evidence does not explain why prices sit where they do, and it contains no recent news about any claimed solution. It also does not show how many people hold each view or how well such markets have forecast outcomes.
The dates need care. Kalshi's excerpt lists its end time as "not supplied," yet the saved market data carries a 2045-01-01 end date [1]. The 2045 date appears in the contract window, not as an announced event. Polymarket's listed end time is Jan 1, 2028, 4:59 AM UTC, while its stated deadline is December 31, 2027, 11:59 PM ET [2]. Readers should check each platform's current rules before relying on any date.
References
- Kalshi: Which Millennium Prize Problem will be solved next?. Provider data observed Nov 22, 2026, 6:00 AM UTC.
- Polymarket: Which Millennium Prize Problem will AI solve next?. Provider data observed Nov 22, 2026, 6:00 AM UTC.
Snapshot observed Nov 22, 2026, 12:00 AM UTC. This analysis describes that snapshot and is not a live news update.
What the contract measures
The two platforms ask related but different questions. Kalshi's event asks which Millennium Prize Problem will be solved next, before 2045 [1]. A problem resolves Yes only if it is the first one solved after issuance. A solution counts once a complete proof, disproof, or counterexample is published in a peer-reviewed journal, the Clay Mathematics Institute or another recognized authoritative academic body recognizes it in writing, or at least two Source Agencies report it [1]. A solution can stop counting if it is retracted or found to contain an identified, unaddressed error [1].
Polymarket's event asks which problem an eligible AI lab will solve first, with a deadline of December 31, 2027, 11:59 PM ET [2]. The announcement must come from an eligible lab, present the solution as the lab's work, and cannot be a partial result [2]. Polymarket also lists a separate outcome for no qualifying AI solution by that deadline [2].
Reading the snapshot
The saved snapshot was observed at 00:00:06 UTC on November 22, 2026, and retrieved at 01:07:28 UTC the same day. It records quoted prices at those times. It is not a current quote, an executable price, or a verified probability.
On Kalshi, the Yes prices are 0.875 for Riemann, 0.085 for Hodge, 0.08 for Yang-Mills, 0.025 for P vs NP, and 0.025 for Birch and Swinnerton-Dyer [1]. On Polymarket, the Yes prices are 0.8125 for Riemann, 0.088 for Hodge, 0.0615 for Yang-Mills, 0.0155 for Birch and Swinnerton-Dyer, 0.011 for P versus NP, and 0.033 for no AI solution by the deadline [2]. The contracts differ in deadline, eligible solvers, and resolution wording, so the two platforms should not be averaged or treated as one probability. A price difference between them is not an arbitrage opportunity by itself.
Both samples are complete as saved. Kalshi includes 5 of 5 markets and Polymarket includes 6 of 6. The reported volumes use different units, 11,442 contracts on Kalshi and $540,008 on Polymarket. Volume is not liquidity or a count of traders, and the two figures should not be compared.
The listed close times also differ. Kalshi's market records show Jan 1, 2045, 4:59 AM UTC, though its event excerpt says the listed end time was not supplied [1]. Polymarket lists Jan 1, 2028, 4:59 AM UTC, which follows its December 31, 2027 deadline wording [2]. These are listed close times, not announced event dates.
What remains uncertain
The evidence does not explain why prices sit where they do, how many traders hold these views, or whether any recent news is relevant. It also does not show whether any problem is near resolution. The sample warning states that the saved events do not establish that every real-world outcome is listed, so omitted outcomes should not be assumed to carry any particular price.
References
- Kalshi: Which Millennium Prize Problem will be solved next?. Provider data observed Nov 22, 2026, 12:00 AM UTC.
- Polymarket: Which Millennium Prize Problem will AI solve next?. Provider data observed Nov 22, 2026, 12:00 AM UTC.
AI-generated analysis based on market data. Not financial advice.
Saved Nov 22, 2026 UTCOverview
The Millennium Prize Problems are seven well-known complex mathematical problems selected by the Clay Mathematics Institute in 2000, each carrying a one million US dollar prize for the first correct solution [3]. Two PredictPedia markets ask which of these problems will be solved next, but they use different rules and different time windows, so they should be read separately. The Kalshi contract asks which Millennium Prize Problem will be solved next before 2045. It resolves to Yes if the first problem solved after issuance is one of the markets listed for the event, with the listed market end date shown as 2045-01-01T04:59:00Z in the market data [1]. A problem counts as solved at the earliest of three conditions: a complete proof, disproof, or counterexample addressing the problem as officially formulated is published in a peer-reviewed journal; the Clay Mathematics Institute or another recognized authoritative academic body officially recognizes in writing that the problem has been resolved by reference to a specific written work; or at least two named Source Agencies report that the problem has been resolved by reference to a specific publicly available written work [1]. The contract also specifies that if several problems share the same solution time, the outcome is divided among the tied problems, and that a qualifying solution stops counting if it is formally retracted, withdrawn, repudiated, rescinded, or found to contain an identified and unaddressed error before expiration [1]. The Polymarket contract is narrower. It asks which Millennium Prize Problem an AI lab will solve first, with a deadline of December 31, 2027, 11:59 PM ET, and a listed end time of 2028-01-01T04:59:00Z [2]. The qualifying problems are the Birch and Swinnerton-Dyer Conjecture, the Hodge Conjecture, P versus NP, the Riemann Hypothesis, and the Yang-Mills existence and mass gap problem. The Navier-Stokes existence and smoothness problem is excluded [2]. A qualifying announcement must come from an eligible AI lab or its official representative, must state that the problem has been solved, and must present the solution as the lab's own work, alone or jointly with outside researchers. Partial results do not qualify, and no confirmation from the Clay Mathematics Institute or another organization is required [2]. If no eligible lab announces a qualifying solution by the deadline, the market resolves to a no-solution outcome [2]. The two contracts also differ in what counts as a solver. Kalshi does not restrict the solution to AI work, while Polymarket restricts qualifying announcements to a long list of named AI labs and to any other company with a model on the MMLU-pro leaderboard at the time of the announcement [1][2]. The market snapshot lists prices for each contract outcome at the time it was observed, and those prices are not executable quotes or verified probabilities. Each platform's reported volume is in its own units and should not be compared across platforms.
Historical Context
The Clay Mathematics Institute selected the seven Millennium Prize Problems in 2000 [3]. The Wikipedia article on the Navier-Stokes priority controversy reports that on 8 September 2026 OpenAI proposed a solution to the Navier-Stokes existence and smoothness problem, and that a mathematician's statement alleged that related Euler equation work had been leaked and may have inspired the prompts used [4]. That article describes the reported dispute, but its retrieval does not establish whether the claimed solution has been accepted under either contract's rules, and the supplied evidence does not give that status.
Why It Matters
Both contracts measure whether a Millennium Prize Problem is solved within a defined window, but they define solved differently. The Kalshi contract uses a standard based on peer-reviewed publication, written recognition by the Clay Mathematics Institute or another recognized academic body, or reports from two Source Agencies, and it allows a solution to be withdrawn if it is retracted or found to contain an unaddressed error [1]. The Polymarket contract instead measures whether an eligible AI lab announces a qualifying solution by December 31, 2027, and it requires the announcement to claim the solution as the lab's work [2]. Because the two contracts measure different events over different periods, their prices answer different questions and should not be treated as directly interchangeable.
Snapshot context
The latest published provider snapshot for this topic was observed on 2026-11-20 at 15:00 UTC. It contains listings from Kalshi and Polymarket. Market descriptions and listed closing times define the contracts; they do not establish that the real-world outcome has occurred. This article does not include a separately verified news update. [1]
Frequently Asked Questions
What must happen for the Kalshi contract to resolve to Yes?
The first Millennium Prize Problem solved after issuance must be one of the listed problems. A problem counts as solved at the earliest of a peer-reviewed publication of a complete proof, disproof, or counterexample, written recognition by the Clay Mathematics Institute or another recognized academic body, or reports from at least two Source Agencies that cite a specific publicly available written work [1].
Does the Polymarket contract require confirmation from the Clay Mathematics Institute?
No. The contract states that no further confirmation from the Clay Mathematics Institute or any other organization is required once an eligible AI lab makes a qualifying announcement [2].
Is the Navier-Stokes problem included in the Polymarket contract?
No. The contract lists the qualifying problems and states that announcements concerning the Navier-Stokes existence and smoothness problem will not qualify [2]. A Wikipedia article reports OpenAI's proposed Navier-Stokes solution from 8 September 2026 [4], but that report does not change the qualifying list.
What happens if a qualifying solution is later retracted under the Kalshi contract?
A qualifying solution ceases to count if, before expiration, it is formally retracted, withdrawn, repudiated, rescinded, or found to contain an identified and unaddressed error [1].
What happens if no qualifying announcement is made under the Polymarket contract?
If no eligible AI lab announces a solution to a qualifying problem by December 31, 2027, 11:59 PM ET, the market resolves to No solution by Dec 31, 2027 [2].
AI-written background. References identify the evidence used; Wikipedia links under Learn More are additional reading. Prices and provider rules are dated snapshots. This is not financial advice.
Related prediction markets
Prices are catalog snapshots and may have changed. Check each market's rules and observation time.

How many more Millennium Prize Problems will AI solve in 2026?
PM $2.79M · K 136.7K contractsWill OpenAI announce another Millennium Prize solution by December 31, 2027?
Kalshi: Will any AI lab solve a Millennium Prize Problem before 2028?
Related contracts
Will an AI lab announce another Millennium Prize solution by December 31, 2027?
Kalshi: Will any AI lab solve a Millennium Prize Problem before 2028?
Related contracts
Will OpenAI announce another Millennium Prize solution by December 31, 2026?
Kalshi: Will any AI lab solve a Millennium Prize Problem before 2027?
Related contracts
Related or unreviewed contracts may settle differently. Their gap is not an equivalent-contract comparison.
Saved prices; check contract rules.

When will the next Millennium Prize be awarded?
PM $121.6K · K 499.8K contractsWill Anthropic announce a Millennium Prize solution by December 31, 2027?
Kalshi: Will Anthropic announces a complete solution, developed by Anthropic or researchers collaborating directly with Anthropic on that solution, to one of the six Millennium Prize Problems unsolved as of September 5, 2026 before Jan 1, 2028?
Related contracts
Will Anthropic announce a Millennium Prize solution by December 31, 2026?
Kalshi: Will a Millennium Prize be awarded for any unsolved Millennium Prize Problem before Jan 1, 2027?
Related contracts
Will Anthropic announce a Millennium Prize solution by December 31, 2026?
Kalshi: Will Anthropic announce a Millennium Prize solution before Jan 1, 2027?
Related contracts
Related or unreviewed contracts may settle differently. Their gap is not an equivalent-contract comparison.
Saved prices; check contract rules.

When will all Millennium Prize Problems be solved?
PM $552.4K · K 9.9K contractsWill CMI declare a Millennium Prize Problem solved by December 31, 2028?
Kalshi: Will all remaining Millennium Prize Problems be solved before 2029?
Related contracts
Will CMI declare a Millennium Prize Problem solved by December 31, 2027?
Kalshi: Will all remaining Millennium Prize Problems be solved before 2028?
Related contracts
Related or unreviewed contracts may settle differently. Their gap is not an equivalent-contract comparison.
Saved prices; check contract rules.

Which math problems will AI solve in 2026?
PM $295.0K
Will Google announce a Millennium Prize solution?
K 11.9K contracts
What will CrowdStrike say during their next earnings call?
K 2.7K contractsMarket Insights
Price gaps of 2 pp+
0
Same-question pairs
0
No outcomes on this page have been reviewed as the same question on both platforms yet.
Price differences do not establish arbitrage. Check each pair's review, current order prices, fees and settlement rules.
