About PhalaCloud

Phala Cloud provides confidential infrastructure for open models, private inference, and verifiable AI workloads.

Phala helps teams run LLMs, agents, and GPU workloads inside hardware-backed trusted execution environments, keeping prompts, data, and runtime secrets private while preserving proof of what ran.

On Hugging Face, we publish model artifacts, benchmarks, datasets, and research resources for confidential AI workflows.

Research

• Dstack: A Zero Trust Framework for Confidential Containers
• Confidential Computing on NVIDIA Hopper GPUs: A Performance Benchmark Study
• Implement Kubernetes Pod-Level Remote Attestation for Confidential Workloads on dstack
• LegalCiteBench: Evaluating Citation Reliability in Legal Language Models
• PASS: A Provenanced Access Subaccount System for Blockchain Wallets
• Persistent BitTorrent Trackers

Hub resources

• Model: PhalaCloud/GLM-5.2-W4AFP8
• Legal citation leaderboard: PhalaCloud/legal-citation-leaderboard
• Legal citation datasets: PhalaCloud datasets
• Cloud: cloud.phala.com
• Trust Center: trust.phala.com
• GitHub: github.com/Phala-Network