List models

Lists the model IDs the gateway can serve right now, in the OpenAI model-list format. The list comes from the gateway's own catalog, not from the model server. Each entry carries the ID you put in model; prices, context length and capabilities are not part of this response.

Prices, context length and capabilities per model are on the Models page.

GET/models

URL

https://api.bricksum.com/v1/models

Request

bash

curl https://api.bricksum.com/v1/models \
  -H "Authorization: Bearer $KILN_API_KEY"

python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.bricksum.com/v1",
    api_key=os.environ["KILN_API_KEY"],
)

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

javascript

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.bricksum.com/v1",
  apiKey: process.env.KILN_API_KEY,
});

const list = await client.models.list();
for (const model of list.data) console.log(model.id);

Response

200 OK

{
  "object": "list",
  "data": [
    {
      "id": "qwen3-32b",
      "object": "model",
      "created": 1710000000,
      "owned_by": "bricksum"
    }
  ]
}

405 Method Not Allowed

{
  "error": {
    "message": "Method POST is not allowed on /v1/models (use GET).",
    "type": "invalid_request_error",
    "param": "",
    "code": "method_not_allowed"
  }
}

Authentication

Authentication is the same on every endpoint the gateway fronts, and is documented on the API Reference page.

Request

Support status

supported: The gateway handles it, or a published measurement shows it works.
passed through: Forwarded to the model server; not verified by Kiln.
ignored: Accepted with 200, but not applied. The rest of the response is normal.
not supported: Rejected with an error, or not working on the models named in the description.

query stringstringignored
The endpoint takes no parameters. A query string is accepted and has no effect on the response. In the fallback described under Response body, it is not forwarded to the model server.
request bodynoneignored
A body sent with GET is not read. Send the request without one.

Only the authentication headers change the outcome, and they are the ones every endpoint takes. There is nothing else to send.

Response body

Field	Type	Description
object	string	The literal "list".
data[].id	string	Model ID, for example qwen3-32b. Put this value in the model field of an inference request.
data[].object	string	The literal "model".
data[].created	integer	Unix seconds, stamped when the gateway builds the response. Every entry of one response carries the same value, and the value moves between calls. It does not record when the model was added.
data[].owned_by	string	The literal "bricksum".

Errors

Errors from this endpoint use the OpenAI envelope: {"error": {"message", "type", "param", "code"}}. Rows with an empty error.code cell are returned by the gateway before the request reaches the endpoint, and use the body shown under the table.

Every status code is listed on the Errors page.

Status · error.code	Cause	What to do
405method_not_allowed	The method is not GET.	Send the request with GET. The response carries an Allow: GET header.
502	Fallback only: the catalog was empty and the model server could not be reached. The body is the plain text upstream error, not JSON.	Retry after a short wait.
503	Fallback only: the catalog was empty and no model server could be picked. The body is the plain text no upstream available, not JSON.	Retry after a short wait, and check the status page.

401, 402, 403 and 429 come from the gateway before the request reaches this endpoint. They are listed once on the API Reference page.

Authentication, credit, budget and rate-limit (requests/min, tokens/min) errors are returned by the gateway before the request reaches this endpoint. Their body carries error.type instead of error.code:

json

{
  "error": {
    "message": "invalid API key",
    "type": "authorization"
  }
}

Limits

Billing: not billed. The gateway publishes no usage event for this endpoint, and the response has no X-Neocloud-Generation-Id header.
Contents: the list holds the models the gateway serves. Deprecated and coming-soon models are left out, so an ID that is missing here is rejected with 404 (model_not_found) by the inference endpoints.
Empty catalog: when the gateway can name no servable model, it proxies the request to a model server and returns that server's own response. The IDs in that body are the model server's, so they can differ from the IDs the gateway serves.
No single-model lookup: GET /models/{model} is not served. The gateway sends that path to a service with no handler for it, so it answers a plain-text 404 rather than an error object. Read one model's entry from the list above.

Headers and retry guidance for 429 are on the Rate limits page.