We are currently observing unprecedented query volume and high concurrent traffic across our DeepSeek V4 Flash model infrastructure.
Impact:
* Users may experience longer time-to-first-token (TTFT) and slower token streaming responses.
* A small percentage of requests may encounter intermittent gateway timeouts or rate-limiting delays.
* Local editor workflows, file operations, and terminal execution remain completely unaffected.
Workaround:
If you require immediate high-speed generation, we recommend temporarily switching to DeepSeek V4 Pro or our fallback agent models in the AI model picker dropdown.
Next Steps:
Our team is actively monitoring connection queues and scaling proxy capacity alongside upstream inference pools to normalize latency. We will provide our next update within 30 minutes or as soon as significant progress is made.
