We are pleased to share that Qwen3.8-27B is now accessible on Phala. The model is currently being hosted inside a TDX-attested GPU TEE. This integration brings you private and verifiable multimodal inference for a wide array of applications. Whether you are focusing on video, images, research, agents, or coding, this setup is ready to assist you. You can find the model at https://phala.com/models/qwen/qwen3.8-27b