DeepSeek routes V4-Pro requests to V4.1-Flash

DeepSeek's API documentation says requests to deepseek-v4-pro are routed to V4.1-Flash from September 14 until V4.1-Pro is released, while the existing endpoint remains available. Why it matters: A stable endpoint can hide a material model change. Outputs, tool use, latency, and spend may change without any integration change.

Try this: Save a small set of representative prompts and tool calls, run them before and after the route change, and compare output quality, failures, latency, and token cost before changing limits or budgets.