Which company has the best AI model end of September?
Anthropic is still the front-runner, but the current price looks more bullish than the underlying uncertainty justifies. I lean yes because Claude remains in the top tier, though a late flip is very plausible.
Analysis
Anthropic enters the final month with real strength, not just hype. The available context suggests Claude Opus 5 and related Claude variants remain near the top on coding, long-context, and broad frontier quality, and that is exactly the type of performance profile that tends to do well on an overall arena leaderboard. If the leaderboard were frozen today, Anthropic would have a credible claim to first place, and the market clearly recognizes that with an extremely aggressive yes price. That said, a favorite is not the same thing as a lock, especially when the underlying ranking is based on a live public leaderboard that can move quickly.
The main reason to temper the yes probability is competitive instability. OpenAI appears close enough to challenge, and some recent comparisons have already put GPT-5.6 Sol ahead of Claude on certain composite or agentic measures. In a market like this, even a modest performance gain, a prompt-tuning update, or a new release can be enough to reorder the top spot. The fact that the top position has reportedly changed hands every few weeks is a strong warning sign against treating Anthropic as near-certain; one month is long enough for a rival lab to overtake, especially if it chooses to push a flagship update before the end-of-September check.
The market price of 91% for yes looks too high relative to that volatility. It implies Anthropic is overwhelmingly likely to stay first, but the evidence more naturally supports a solid favorite rather than a near certainty. My assessment leans yes because Anthropic seems well positioned and already has broad support in both benchmarks and trader sentiment, but I would still assign a meaningful chance that OpenAI or another lab finishes the month at the top. The decisive factor will be whether any competitor lands a late-cycle release strong enough to surpass Claude exactly at the September 30 noon ET snapshot.
Arguments
For
- Arguments for Yes: Claude variants are already among the strongest models on the benchmark types that matter for arena-style ranking.
- Arguments for Yes: If no major rival ships a clearly superior model before September 30, Anthropic has a good chance to keep the lead.
Against
- Arguments against Yes: The top spot is contested, and GPT-5.6 Sol has already shown enough strength to threaten Anthropic on some comparisons.
- Arguments against Yes: A single late September improvement from a competitor could be enough to change the first-place company on the final check.
Key drivers
- Claude appears to be near the top already, which gives Anthropic a strong starting position.
- The arena leaderboard is volatile, so the final ranking can still shift in a short time window.
- A late release or update from a rival lab could change the top spot right before the check.
- Trader sentiment is strongly pro-Anthropic, which signals real but possibly overstated confidence.
Risk factors
- OpenAI could push GPT-5.6 Sol or another model back into first place before the resolution time.
- A leaderboard update, tie-break, or score revision could overturn a narrow Anthropic lead.
- Anthropic may stay competitive without actually finishing first on the specific overall ranking used for resolution.
- The market may be pricing in current leadership too aggressively and underestimating late movement.
Scenarios
Best case
Anthropic keeps the lead through the end of September, or releases a stronger Claude update that widens its advantage before the leaderboard snapshot.
Most likely
Anthropic remains in the leading tier and has a slightly better than even chance of still being first, but the final outcome stays vulnerable to a late competitive swing.
Worst case
A rival model, most plausibly from OpenAI, overtakes Claude in the final days and Anthropic finishes second at the September 30 check.
More from this day
- pop culturePolymarket3mo
Top Spotify Artist 2026
AI36%MKT88%Edge-52HypedBad Bunny is still a real contender because of his global fanbase and recent streaming strength, but the available evidence does not support treating him as close to certain for 2026 #1. I estimate his chance well below the market price because other mega-artists still look better positioned to finish the year on top.
- EconomicsKalshi7y
US real GDP growth in 2033?
AI58%MKT12%Edge+46Hidden GemThe most likely outcome is that US real GDP growth in 2033 lands in the mid-range, with a meaningful but not dominant chance of a recessionary or very weak-growth year. My independent view is somewhat less bearish than the market on the top-range scenarios, and somewhat more supportive of moderate growth outcomes.
- FinancialsKalshi13y
Will OpenAI or Anthropic IPO first?
AI58%MKT94%Edge-36HypedAnthropic appears slightly more likely to IPO first than OpenAI, despite the market heavily favoring OpenAI. The current price looks directionally plausible on long-term brand recognition but too confident given the reported timing gap and the fact that both companies are still at the confidential-filing stage.