Features and improvements
October 23, 2026
US (Virginia) region launched
QwenCloud now serves models from US (Virginia) (us-east-1). Workspaces in this region use a service deployment scope: Global or US. The region endpoint is https://us-east-1.maas.qwencloudapi.com. Real-time model calls, Try AI, API keys, workspaces, and billing are supported; Batch, datasets, fine-tuning, evaluation, deployment, and alerts are not available in this region. See Regions and endpoints.October 21, 2026
Multi-region support launched
Multi-region support is now available. QwenCloud supports Singapore, China North 2 (Beijing), Germany (Frankfurt), China (Hong Kong), and Japan (Tokyo). Each region has its own endpoint, API keys, workspaces, models, and prices. Code samples use the Singapore endpointap-southeast-1.maas.qwencloudapi.com. Existing code that uses maas.qwencloudapi.com continues to call Singapore. See Regions and endpoints.May 15, 2026
Model production
The console now supports model production workflows, including dataset management, fine-tuning (SFT/LoRA), and model deployment. You can upload training data, create fine-tuning jobs, and deploy trained models as inference services.Also launched Batch API in the console for submitting and managing batch inference jobs.March 31, 2026