Skip to main content
POST
/
v1alpha1
/
inference
/
gateways
Create gateway
curl --request POST \
  --url https://api.coreweave.com/v1alpha1/inference/gateways \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "name": "<string>",
  "zones": [
    "<string>"
  ],
  "id": "<string>",
  "coreWeaveAuth": {},
  "weightsAndBiasesAuth": {
    "apiKey": "<string>",
    "serverUrl": "<string>",
    "enableUsageReports": true,
    "enableRateLimiting": true
  },
  "endpointConfiguration": {
    "additionalDns": [
      "<string>"
    ]
  },
  "pathBasedRouting": {}
}
'
{
  "gateway": {
    "spec": {
      "id": "<string>",
      "name": "<string>",
      "zones": [
        "<string>"
      ],
      "coreWeaveAuth": {},
      "weightsAndBiasesAuth": {
        "apiKey": "<string>",
        "serverUrl": "<string>",
        "enableUsageReports": true,
        "enableRateLimiting": true
      },
      "organizationId": "<string>",
      "endpointConfiguration": {
        "additionalDns": [
          "<string>"
        ]
      },
      "bodyBasedRouting": {},
      "headerBasedRouting": {
        "headerName": "<string>"
      },
      "pathBasedRouting": {}
    },
    "status": {
      "createdAt": "2023-11-07T05:31:56Z",
      "updatedAt": "2023-11-07T05:31:56Z",
      "conditions": [
        {
          "type": "<string>",
          "lastUpdateTime": "2023-11-07T05:31:56Z",
          "reason": "<string>",
          "message": "<string>",
          "zone": "<string>"
        }
      ],
      "endpoints": [
        "<string>"
      ]
    }
  }
}
Each gateway must be configured with exactly one authentication type (coreWeaveAuth or weightsAndBiasesAuth) and exactly one routing strategy (bodyBasedRouting, headerBasedRouting, or pathBasedRouting). The available zones for zones are returned by GET /v1alpha1/inference/gateways/parameters.
Example request
curl -X POST https://api.coreweave.com/v1alpha1/inference/gateways \
       -H "Content-Type: application/json" \
       -H "Authorization: Bearer {API_ACCESS_TOKEN}" \
       -d @data.json

​
Authentication

  • coreWeaveAuth: validate inference requests using the same CoreWeave API access tokens used to call the management API. No additional configuration is required.
  • weightsAndBiasesAuth: validate inference requests using Weights & Biases SaaS credentials. Optional fields include enableUsageReports and enableRateLimiting. W&B self-hosted is not currently supported.

​
Routing strategies

  • bodyBasedRouting: the model name is read from the model field in the JSON request body. Set apiType: API_TYPE_OPENAI for OpenAI-compatible behavior.
  • pathBasedRouting: the model name is the first segment of the URL path. URL-encode model names that contain special characters.
  • headerBasedRouting: the model name is read from a custom HTTP header you specify in headerName (1–100 characters).

Authorizations

​
Authorization
string
header
default:Bearer {API_ACCESS_TOKEN}
required

CoreWeave API access token sent as a bearer token.

Body

application/json

Request for CreateGateway

​
name
string
required

The human readable name of the gateway

​
zones
string[]
required

The zones to make the gateway available in, limits where deployments associated with the gateway may exist, no zones means all may be used, case insensitive.

​
id
string

The unique identifier of the gateway, UUID format

​
coreWeaveAuth
object

CoreWeave IAM

​
weightsAndBiasesAuth
object

Weights & Biases

​
endpointConfiguration
object

Additional endpoint configuration options

​
bodyBasedRouting
object

Body based routing

​
headerBasedRouting
object

Header based routing

​
pathBasedRouting
object

Path based routing

Response

OK

Response for CreateGateway

​
gateway
object

The gateway that was created

Last modified on May 22, 2026