Which company has best AI model end of August?
Anthropic is the clear favorite, and the current market price already implies a very high probability of holding first place. I still leave a small but meaningful chance for a late OpenAI, Google, or other frontier release to overtake it before the August 31 check.
Analysis
Anthropic enters this market with the strongest current position in the available evidence, and the market itself already prices a very high chance of a Yes outcome at 91.5%. The supplied leaderboard-style sources place Anthropic models at or near the top of text and overall model rankings, including one source that lists Claude Fable 5 as the top model and another that identifies Claude Opus 4.6 thinking as the current overall text leader. That combination suggests Anthropic is not just popular in prediction markets but also competitive on the underlying benchmark the market uses to resolve.
The main argument for Yes is that the resolution criterion is narrow and specific: only the company owning the model with the highest arena rank at the August 31 check matters. If Anthropic remains in first place or even preserves a small lead, it wins regardless of broader perceptions about model quality. The current evidence also suggests Anthropic has a lead across multiple text-oriented views, which increases the chance that it remains on top unless a rival makes a very strong late move. With less than six weeks until resolution, a stable leader in a crowded frontier can still be favored if no major breakthrough lands in the final stretch.
The strongest argument against Yes is that this market is highly exposed to late releases and benchmark volatility. The provided context notes that OpenAI, Google DeepMind, and other labs are close enough on different tasks that a new model could change the ranking quickly, especially if the leaderboard methodology or the exact evaluated model set shifts. Because the market resolves to the single top-ranked company at one point in time, Anthropic does not need to be broadly best across the ecosystem; it only needs to avoid being displaced on that specific leaderboard snapshot. That makes the current favorite status meaningful, but not close to certain.
Arguments
For
- Arguments for Yes: Anthropic currently appears to have the strongest overall position on the relevant text leaderboard and in market pricing.
- Arguments for Yes: The market resolves on a single rank check, and Anthropic only needs to remain first at that moment.
Against
- Arguments against Yes: The field is crowded, and a late OpenAI, Google, or other frontier release could overtake Anthropic.
- Arguments against Yes: Different sources rank top models differently, which shows that the lead is real but not secure.
Key drivers
- Anthropic is already the market leader, and the market price implies a strong expectation that it stays first.
- The resolution depends on a single leaderboard snapshot, so a small edge at the end is enough to win.
- Recent leaderboard sources place Anthropic models at or near the top of the frontier across text tasks.
- Late model releases from OpenAI, Google, or others could still overturn a narrow lead before the August 31 check.
Risk factors
- A late August release from a competing lab could jump ahead on the arena ranking and flip the outcome.
- The frontier is tightly clustered, so small benchmark or score changes can matter more than the headline market favorite status.
Scenarios
Best case
Anthropic preserves or slightly expands its lead through August, and no competing release surpasses it at the final leaderboard check.
Most likely
Anthropic remains very near the top, but the outcome depends on whether any competitor manages a final late-cycle jump; the most likely result is a narrow Anthropic win.
Worst case
A rival lab ships a stronger model late in the month and takes the top spot on the arena leaderboard at the resolution time.
More from this day
- pop culturePolymarketEnded
Harvey Weinstein prison time?
AI3%MKT97%Edge-94HypedWeinstein is far more likely than not to receive some prison time in this matter, because the available reporting indicates he still faces an active New York sentencing process and already has a substantial California sentence on the books. The market’s “No prison time” outcome looks very unlikely unless the remaining New York process collapses in an unexpected way and no custodial sentence is imposed by the deadline.
- pop culturePolymarketEnded
"Spider-Man: Brand New Day" Opening Weekend Box Office
AI93%MKT4%Edge+89Hidden GemThe market strongly favors a debut above $200 million, and the latest tracking is clustered well past that threshold. The main uncertainty is that one older Deadline-based estimate still sits below $200 million, but it now looks like the lower-probability outcome.
- PoliticsKalshi2y
Which agencies will Trump eliminate?
AI86%MKT30%Edge+56Hidden GemI assess a high probability that USAID will be eliminated in the practical, market-relevant sense during Trump’s term. The reporting shows the agency has already been effectively dismantled and folded into the State Department, though there is some legal ambiguity about whether a formal statutory abolition has fully occurred.