Provider pricing

Phala Network AI API pricing

Phala Network is a confidential compute provider founded in 2018 and headquartered in Singapore. Runs LLM inference inside Intel TDX and NVIDIA H100 TEE enclaves with verifiable attestation.

Founded2018
HeadquartersSingapore
Tracked modelsDirectory

Best for

Privacy-critical workloads needing verifiable confidential inference.

Strengths

  • TEE-protected inference on NVIDIA H100 confidential GPUs
  • On-chain attestation of model and prompt integrity
  • OpenAI-compatible confidential endpoints
  • Listed in the TokenCost AI providers directory