Will OpenAI have the best Code Arena | WebDev AI at the end of September 2026?

ResolutionResolved NO

YES payout0¢

NO payout100¢

VenuePolymarket

Conviction at publication

Realized · held to resolution−100%$100 paper stake

Polymarket71¢ → 0¢

Sep 21 forecast

OpenAI's GPT-6 Astra currently leads the Code Arena WebDev leaderboard with a 42-point margin, so it is likely to remain the top model at month-end.

OpenAI's GPT-6 Astra has maintained a clear lead on the Code Arena WebDev leaderboard throughout September 2026, and with only ten days remaining, it is likely to remain the top model at month-end. The key risk is that Anthropic may release an expedited model that outperforms GPT-6 Astra, as suggested by recent reports of internal discussions, and Anthropic's Claude Fable 5.1 already leads the separate SWE-bench Pro benchmark, indicating strong coding ability that could translate to WebDev. However, any new Anthropic model would need to enter the leaderboard and surpass a 42-point gap within a very short time frame. Markets currently treat this outcome as favorable for OpenAI, and the primary consideration behind any difference is whether Anthropic can execute an accelerated release in the final days. Overall, the evidence supports a high likelihood that OpenAI retains the top spot.

Points to YES

OpenAI's GPT-6 Astra holds a consistent lead on the Code Arena WebDev leaderboard, having been ranked #1 on multiple snapshots since its launch.
As of September 21, 2026, GPT-6 Astra leads with 1,800 points, 42 points ahead of Claude Fable 5.1.

Points to NO

Anthropic is considering an expedited model release to counter GPT-6 Astra's lead, with leadership reviewing the timeline.
On September 21, 2026, reports indicated Anthropic leadership is reviewing whether to accelerate its next model iteration.
Anthropic's Claude Fable 5.1 leads the SWE-bench Pro leaderboard, demonstrating strong coding capabilities that could translate to WebDev performance.
As of September 21, 2026, Claude Fable 5.1 has an 81.2% resolution rate on SWE-bench Pro, while OpenAI has not published verified figures.
On September 5, OpenAI and Anthropic were tied for rank 1-2 despite OpenAI's raw-point lead, indicating that small gaps can produce ranking ties.
On September 5, 2026, OpenAI's GPT-6 Astra Max and Anthropic's Claude Fable 5.1 Max shared rank 1-2 due to overlapping statistical score ranges despite a 35-point raw lead.

Research and information only. Not financial advice or a recommendation to trade. Markets involve risk, including loss of capital. Review venue rules and eligibility before trading.