With Private AI Proxy
- Install Private AI Proxy, choose a Phala profile, sign in, and turn on Protected.
- On the Agents screen, connect Claude Code.
- Start
claude.
Without the Proxy
Create a key under Private AI → API Keys in the Phala dashboard. Set the base URL to the host without/v1; Claude Code appends /v1/messages itself.
ANTHROPIC_AUTH_TOKEN sends the key as a bearer token; clearing ANTHROPIC_API_KEY stops an
existing Anthropic key from adding a conflicting x-api-key header.
Phala model IDs do not appear in the /model picker; set the model with ANTHROPIC_MODEL or
/model z-ai/glm-5.3. The alias variables send plan mode and background work to a Phala model.
This path does not verify attestation locally; use Private AI Proxy or the TypeScript ACI Verifier. Every model on this endpoint runs in a TEE.
Check the Connection
Run/status in Claude Code. With the Proxy, the base URL is http://127.0.0.1:4180; without it,
https://inference.phala.com. Confirm the bearer credential source and the selected model.

