List Models

List available Confidential AI models, modalities, context windows, and pricing metadata.

Endpoints

GET https://inference.phala.com/v1/models

Use the live model catalog before hardcoding model IDs. The catalog returns model IDs, context windows, pricing, serving metadata, modalities, supported parameters, and whether a model can be served confidentially.

Query Parameters

zdr · string

Set to true to return only models that can be served under zero data retention, meaning the upstream serving route does not retain prompt or completion content.

A model is included when at least one route that can serve it operates under zero data retention. Listing a model here means a request for it with provider: {"zdr": true} can be routed. See Zero Data Retention.

true is the only accepted value. Any other value, including false, returns 400. Omit the parameter to list the full catalog.

curl "https://inference.phala.com/v1/models?zdr=true" \
  -H "Authorization: Bearer <API_KEY>"

tee · string

Set to true to return only models that are offered confidentially, meaning the model object carries is_tee: true.

This filter reports a model’s own capability. It does not promise that a given request will reach an attested provider, which depends on the deployment serving it. To require attestation for a request, send provider: {"aci_verified": true}. See Attested Routing.

true is the only accepted value. Any other value, including false, returns 400. Omit the parameter to list the full catalog.

curl "https://inference.phala.com/v1/models?tee=true" \
  -H "Authorization: Bearer <API_KEY>"

Both filters compose. Combining them returns models that satisfy each constraint.

Response entries have the same shape as the unfiltered catalog. There is no zdr field on a model object: the filter is the way to read this property.

An invalid value returns 400:

{
  "error": {
    "message": "Invalid 'zdr' query parameter. The only accepted value is 'true'; omit it for unfiltered results.",
    "type": "invalid_request_error"
  }
}

Examples

All Models

curl https://inference.phala.com/v1/models \
  -H "Authorization: Bearer <API_KEY>"

from openai import OpenAI

client = OpenAI(
    api_key="<API_KEY>",
    base_url="https://inference.phala.com/v1",
)

models = client.models.list()
for model in models.data:
    print(model.id)

Response

{
  "data": [
    {
      "id": "phala/qwen3.5-27b",
      "name": "Qwen3.5 27B",
      "created": 1677652288,
      "is_tee": true,
      "description": "Qwen model running through Phala GPU TEE infrastructure",
      "context_length": 262144,
      "max_output_length": 262144,
      "pricing": {
        "prompt": "0.00000030",
        "completion": "0.00000240"
      },
      "providers": ["phala"],
      "input_modalities": ["text"],
      "output_modalities": ["text"],
      "supported_parameters": ["max_tokens", "temperature", "tools", "tool_choice", "response_format"],
      "metadata": {}
    }
  ]
}

Model Object Fields

Field	Description
id	Model identifier for API calls
name	Human-readable model name
is_tee	true if the model can be served confidentially by a verified TEE provider
description	Model or provider description
context_length	Maximum context window
max_output_length	Maximum output length
pricing.prompt	Input token price per token; multiply by 1,000,000 for per-million-token pricing
pricing.completion	Output token price per token; multiply by 1,000,000 for per-million-token pricing
providers	Serving routes available for the model.
input_modalities	Supported input types, such as text or image
output_modalities	Supported output types, such as text or embeddings
supported_parameters	Request parameters accepted by the model

Find Verifiable TEE Models

Filter for models that can be served confidentially:

curl "https://inference.phala.com/v1/models?tee=true" \
  -H "Authorization: Bearer <API_KEY>" | \
  jq -r '.data[].id'

is_tee: true means the model can be served confidentially. To require attestation for a specific request rather than reading the catalog flag, use Attested Routing. The receipt remains the per-response proof: read the x-receipt-id header, then verify it with Attestation Report and Get Receipt.

Find Zero-Data-Retention Models

is_tee and zdr describe different properties. is_tee says a model can run inside a verified TEE, which constrains who can read memory during execution. Zero data retention says the upstream serving route does not keep prompt or completion content after serving the request. Filter for it with the query parameter rather than inferring it from is_tee:

curl -s "https://inference.phala.com/v1/models?zdr=true" \
  -H "Authorization: Bearer <API_KEY>" | \
  jq -r '.data[].id'

To route a request under the same constraint, send provider: {"zdr": true} in the request body. See Zero Data Retention.