DeepSeek’s rapid release cadence, including the August 2026 general availability of V4 Pro with upgraded agent capabilities and OpenAI Responses API support, followed by the September launch of the smaller V4.1 Flash model, has set expectations for an imminent V4.1 Pro successor. The company temporarily routes V4 Pro API traffic to V4.1 Flash at lower cost, signaling that the next flagship mixture-of-experts model remains in active development amid intense competition from other Chinese labs releasing models every few weeks. Traders watch for official announcements on benchmarks, context-window improvements, Ascend-chip optimizations, and pricing changes, as these verified milestones typically shift market-implied odds more than unconfirmed rumors.