Traders assign just 2.5 percent odds to any Chinese AI model claiming the top spot on major leaderboards by June 30, reflecting the continued performance edge held by U.S.-developed systems across core benchmarks. Current gaps in training compute scale, multimodal capabilities, and inference efficiency remain sizable, with recent Western releases reinforcing those leads through incremental upgrades rather than dramatic leaps. Chinese teams have advanced open-source and domestically optimized models in the past month, yet none have closed the distance enough to threaten the frontrunners. A sudden high-impact release demonstrating superior results on standardized evaluations could still shift sentiment, though the short remaining window limits the time available for such a turnaround to register and stabilize.