Skip to content

Encrypted inference

API

Explore the OpenAI-compatible endpoint and private transport before creating keys, adding credits, and monitoring usage.

Sign in to add 500 credits
OverviewReference

Garlic API

Build private AI with the tools you already use.

Point any OpenAI SDK at one base URL. Choose standard JSON, an explicit TEE model, or E2EE by changing only the model name.

curl https://garlic.ai/v1/chat/completions \
  -H "Authorization: Bearer $GARLIC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "tee-privatemode/gpt-oss-120b",
    "messages": [{"role": "user", "content": "Hello, private world!"}]
  }'

Model IDs are mode-provider/model, always in full, such as tee-privatemode/gpt-oss-120b; a bare name is rejected, and only an explicit passthrough- ID ever leaves the confidential path. Browser chat uses a separate browser-side E2EE transport.

What sign-in unlocks

API keysCreate and revoke personal keys
BillingAdd credits without overages
UsageMonitor requests and token spend
TransportStandard · TEE

Sign in when you’re ready. The endpoint and examples stay public. See per-token pricing.

Using an agent?

The same installer configures the provider and default model for OpenClaw and Hermes.