Skip to content

Model Availability

The set of models you can call is determined by your account configuration, not hard-coded into the API. Two requests with different keys may see different models.

How availability is resolved

Each request is associated with a key pool, which defines the models and capabilities available to the caller. You can target a specific pool with the X-Key-Pool-ID header; otherwise the pool bound to your key is used.

Header Required Description
Authorization Yes Bearer <api-key> — primary authentication for all endpoints.
X-Key-Pool-ID Recommended Determines which models and capabilities are available to the request.
X-User-ID No End-user identifier for routing and logging.
X-API-Key-ID No Specific API key identifier.

When using the Anthropic Messages format, you may authenticate with the x-api-key header instead of Authorization (Anthropic SDK compatible). The router accepts either.

Discover available models

List the models available to your key with a single call — this reads your configuration and does not incur a model invocation:

curl https://llm.siraya.ai/v1/models \
  -H "Authorization: Bearer <API_KEY>"

from openai import OpenAI

client = OpenAI(base_url="https://llm.siraya.ai/v1", api_key="<API_KEY>")
for m in client.models.list().data:
    print(m.id)

See the Models catalog for the full list of model families and the capabilities each supports.