Next Claude Opus Model: Humanity’s Last Exam Debut?
The evidence strongly favors a Yes outcome, since the latest Claude Opus release already appears to clear 45% on Humanity’s Last Exam by a wide margin. The main uncertainty is not whether it beats the threshold, but which published score the market will use, and all currently cited variants are still comfortably above 45%.
Analysis
The current setup points heavily toward a Yes result because the most recent Claude Opus flagship already debuted with HLE performance far above the 45% bar. The recent published figures reported for Claude Opus 5.5 are clustered in the 55% to high-60% range depending on the evaluation setup, which is a substantial cushion above the threshold. That means even if the market is meant to resolve off a conservative or lower-end published score, the model still appears to clear the requirement with room to spare.
The biggest analytical issue is not the underlying capability level, but the resolution mechanics. The market depends on the next Claude Opus model newly added to the Humanity’s Last Exam results page, and the displayed HLE Accuracy at the relevant time is what counts. Recent context suggests there may be multiple benchmark presentations, such as with tools versus without tools, or different leaderboard conventions. However, every cited version in the provided news context still lands above 45%, so the ambiguity changes the exact margin, not the direction of the outcome.
Market pricing is consistent with that interpretation. A Yes price of 0.9325 implies traders already think the threshold is very likely to be exceeded, and that seems reasonable given the recent benchmark data and Anthropic’s apparent trajectory on this benchmark. The remaining No probability is mostly a residual for edge cases such as a future Claude Opus release unexpectedly underperforming, a delayed or missing site update, or a resolution quirk that prevents the model from qualifying as intended. Those are real risks, but they are not strong enough to overcome the current evidence.
Overall, the market looks close to a near-lock unless there is a surprising change in how the next Opus model is scored or displayed. The latest available information suggests Anthropic is already operating well above the threshold, and the threshold itself is not especially demanding relative to the published results. That makes Yes the clear favorite, with only a small tail risk from benchmark-format confusion or an unforeseen drop in the next model’s debut score.
Arguments
For
- Arguments for Yes: The latest Claude Opus release has already been reported in the mid-50s to high-60s on HLE, which is comfortably above 45%.
- Arguments for Yes: Even the lower-end third-party figures in the provided context remain above the threshold, reducing the chance that resolution depends on a borderline score.
Against
- Arguments against Yes: The market depends on the exact displayed result card, so a formatting or benchmark-label mismatch could create resolution uncertainty.
- Arguments against Yes: A future Opus debut could theoretically underperform relative to the latest release, although that seems unlikely given the current trend.
Key drivers
- Recent Claude Opus HLE results are already well above 45%, leaving a wide safety margin.
- The market resolves off the displayed HLE Accuracy, and all currently cited published variants are still above the cutoff.
Risk factors
- The next qualifying Opus model could use a different benchmark setup or display convention than the one traders expect.
- A site update delay, removal, or unusual resolution edge case could complicate qualification even if the score itself is high.
Scenarios
Best case
The next Claude Opus model appears with an HLE score clearly above 45%, likely in the 50s or 60s, and the result is visible on the site at the required time with no ambiguity.
Most likely
A new Claude Opus model appears on the results page with an HLE score well above 45%, and the market resolves Yes with little controversy.
Worst case
The qualifying model is delayed, removed, or shown in a way that prevents proper qualification, or the next Opus debut unexpectedly lands below 45% on the displayed score.
More from this day
- PoliticsKalshi23d
Will 3 mg THC beverages be federally legal to sell?
AI9%MKT94%Edge-85HypedA 3 mg THC beverage is very unlikely to remain federally legal to sell on Nov. 16, 2026 under the newly reported hemp rules. I put the Yes probability at 9%, with the market appearing heavily overpricing legalization persistence.
- pop culturePolymarketEnded
"Forgotten Island" Opening Weekend Box Office
AI28%MKT80%Edge-52HypedThe pre-release tracking points to an opening weekend comfortably above $13 million, so I think the chance of a sub-$13 million result is fairly low. The market looks close to even, but the industry forecasts lean enough to the upside that I would favor No.
- PoliticsKalshi2mo
When will a catastrophic AI duty of care become U.S. law?
AI58%MKT8%Edge+50Hidden GemI put the odds of Yes meaningfully above the market because several state laws now create AI safety obligations that plausibly fit a catastrophic-risk duty of care before the deadline. The main reason not to go much higher is that the exact contract wording may be read narrowly as federal law only, which would leave the outcome much less likely.