OpenAI holds a commanding 76.8% implied probability as the leader in Code Arena WebDev AI models by end of September, reflecting trader consensus on its sustained edge in large language model coding performance. Recent model iterations have demonstrated stronger results on web-specific tasks such as frontend component generation, API integration, and responsive design benchmarks compared with rivals. Anthropic at 23.5% captures the main alternative view, driven by its focus on reliable, safety-tuned outputs that appeal in enterprise web development scenarios. Other listed companies remain near zero or at 50% in multi-outcome formats, underscoring limited credible evidence of competitive breakthroughs in the narrow 10-day window. Key near-term catalysts include any final benchmark releases or capability demonstrations before month-end that could shift the tight resolution criteria.