Which company has the best AI model end of November?
Anthropic remains a serious contender, but the current public leaderboard leadership appears to be with Google, and OpenAI is also close enough to make this a three-way race. I would put Anthropic’s chance of finishing first at about 34%.
Analysis
The central issue is that this market resolves to the company whose model is ranked first on the Arena overall text leaderboard at the end of November, not to whichever company has the strongest reputation or the best model on some narrower benchmark. On the latest signals, Anthropic is clearly in the top tier, but it is not the clear front-runner in the exact ranking that matters. Google’s Gemini 4 Argon High is currently described as sitting at the top of the overall leaderboard, which means Anthropic would need to either overtake it or benefit from Google slipping before the check date. That makes Anthropic live, but not favored, in the strict sense of this market rule.
Arguments for Yes are still meaningful because Anthropic has repeatedly shown strength in areas that tend to matter for Arena-style preference rankings, including helpfulness, coding quality, and agentic task performance. The market also appears to value Anthropic highly in related but narrower venues, which suggests traders believe Claude models remain competitive enough to surge into first place if product quality, prompt handling, or release timing breaks its way. If Anthropic ships an especially strong update before the end of November, it could plausibly reclaim the top spot, especially in a leaderboard environment where small preference shifts can change rank quickly.
Arguments against Yes are stronger at the moment because the most direct evidence points to Google holding the current lead, while OpenAI remains capable of disrupting the order as well. This is not a situation where Anthropic only needs to defend a comfortable first place; it is trying to climb past at least one, and possibly two, major rivals in a dynamic ranking system. The broader market data also suggests uncertainty about what “best” means across tasks, and that matters because Arena overall reflects aggregate preference rather than one specialized strength. With time remaining until the end of November, Anthropic has a path to win, but the current position and competitive pressure make a No outcome slightly more likely than Yes.
Arguments
For
- Anthropic is still widely viewed as one of the strongest model families, so it has a plausible route back to number one.
- If Arena voters continue to favor Claude on nuanced, high-quality responses, Anthropic could outperform its competitors by the resolution date.
Against
- The latest public leaderboard signal points to Google being ahead right now, which means Anthropic is not currently in the winning position.
- OpenAI’s recent model launches keep the competitive field crowded, reducing the chance that Anthropic can stay or move into first place.
Key drivers
- Anthropic needs to overtake the current leaderboard leader rather than merely hold a small margin.
- Arena rankings can move quickly after major model releases or updates, especially among frontier labs.
- Google and OpenAI both have credible paths to remain ahead of Anthropic through November.
Risk factors
- A new Google release or improvement could widen the gap and lock Anthropic out of first place.
- OpenAI could leapfrog both rivals if its newer models gain broader user preference on Arena.
- The market is sensitive to exact check-time ranking, so even a brief late-November slip would decide the event.
Scenarios
Best case
Anthropic releases or benefits from a model update that improves overall user preference enough to surpass Google and stay ahead through the November 30 check.
Most likely
Anthropic remains highly competitive but finishes behind either Google or OpenAI, with the exact order depending on late model updates and leaderboard volatility.
Worst case
Google maintains the top spot or OpenAI overtakes both, leaving Anthropic in second place or lower at resolution.
More from this day
- pop culturePolymarketEnded
#2 Spotify song in the US this week? (October 9)
AI14%MKT87%Edge-73HypedChoosin' Texas is still an enormous hit, but the strongest recent signals point to another song taking the Spotify US #2 spot this week. I see this as a low-probability Yes outcome despite the song’s overall dominance in broader music charts.
- pop culturePolymarket21d
"Resident Evil" total domestic gross by October 31?
AI84%MKT42%Edge+42Hidden GemResident Evil is already above 100 million domestic, but the remaining runway to October 31 looks too short for it to add the roughly 45 million needed to reach 150 million. The current pace and normal post-opening decay strongly favor a finish below the threshold.
- SportsKalshiEnded
San Jose St. vs Hawai'i: Total Points
AI92%MKT54%Edge+38Hidden GemI think Over 43.5 is very likely to cash, with the pregame total and live game state both pointing to a finish comfortably above the number. My independent estimate is about 92% Yes.