Content is transient
Prompts and completions are processed to provide inference and are not persisted by our inference gateway.
Independent inference provider · China
Clear terms for
reliable AI inference.
We provide OpenAI-compatible text inference with transparent data practices and straightforward service terms.
No model training
We do not use customer prompts or completions to train models.
Operational metadata only
We retain limited request and usage metadata for billing, abuse prevention, and reliability.