Update AI configuration

PUT

/api/ai/config

curl --request PUT \
  --url https://gateway.example.com/api/ai/config \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{ "enabled": true, "providerType": "openai_compatible", "providerUrl": "", "endpointMode": "auto", "supportsImages": true, "apiKey": "example", "model": "example", "gatewayInferenceModel": "example", "gatewayInferenceAllowUserModelSelection": true, "allowUserReasoningEffortSelection": true, "customSystemPrompt": "example", "rateLimitMax": 1, "rateLimitWindowSeconds": 1, "maxToolRounds": 1, "maxContextTokens": 1, "maxCompletionTokens": 1, "maxTokensField": "max_tokens", "reasoningEffort": "low", "disabledTools": [ "example" ], "webSearchApiKey": "example", "webSearchProvider": "tavily", "webSearchBaseUrl": "", "sandboxEnabled": true, "sandboxDefaultTier": "low" }'

• Your Gateway instance

Authorizations

• bearerAuth

Request Body^required

JSON request body

Media typeapplication/json

object

enabled

boolean

providerType

string

Allowed values: openai_compatible gateway_inference

providerUrl

Any of:

string

string format: uri

string

string

Allowed values: ""

endpointMode

string

Allowed values: auto chat_completions responses

supportsImages

boolean

apiKey

string

model

string

gatewayInferenceModel

string

<= 255 characters

gatewayInferenceAllowUserModelSelection

boolean

allowUserReasoningEffortSelection

boolean

customSystemPrompt

string

rateLimitMax

integer

>= 1 <= 1000

rateLimitWindowSeconds

integer

>= 10 <= 3600

maxToolRounds

integer

>= 1 <= 50

maxContextTokens

integer

>= 4000 <= 1000000

maxCompletionTokens

integer

>= 256 <= 128000

maxTokensField

string

Allowed values: max_tokens max_completion_tokens

reasoningEffort

string

Allowed values: low medium high none

disabledTools

Array<string>

webSearchApiKey

string

webSearchProvider

string

Allowed values: tavily brave serper searxng exa

webSearchBaseUrl

Any of:

string

string format: uri

string

string

Allowed values: ""

sandboxEnabled

boolean

sandboxDefaultTier

string

Allowed values: low medium high

Responses

200

Successful response

Media typeapplication/json

object

data

required

Resource object. Exact fields depend on the endpoint.

object

id

string

name

string

key

additional properties

Example

{
  "data": {
    "enabled": true,
    "providerUrl": "https://api.openai.com/v1",
    "model": "gpt-4.1"
  }
}

400

Bad request

Media typeapplication/json

object

code

required

Error code

string

message

required

Human-readable error message

string

details

Additional error details

Example

{
  "code": "VALIDATION_ERROR",
  "message": "Request validation failed",
  "details": [
    {
      "path": [
        "name"
      ],
      "message": "Required"
    }
  ]
}

401

Authentication required

Media typeapplication/json

object

code

required

Error code

string

message

required

Human-readable error message

string

details

Additional error details

Example

{
  "code": "UNAUTHORIZED",
  "message": "Authentication required"
}

403

Forbidden

Media typeapplication/json

object

code

required

Error code

string

message

required

Human-readable error message

string

details

Additional error details

Example

{
  "code": "FORBIDDEN",
  "message": "Insufficient permissions"
}

404

Not found

Media typeapplication/json

object

code

required

Error code

string

message

required

Human-readable error message

string

details

Additional error details

Example

{
  "code": "NOT_FOUND",
  "message": "Resource not found"
}

409

Conflict

Media typeapplication/json

object

code

required

Error code

string

message

required

Human-readable error message

string

details

Additional error details

Example

{
  "code": "CONFLICT",
  "message": "Resource already exists"
}

422

Unprocessable entity

Media typeapplication/json

object

code

required

Error code

string

message

required

Human-readable error message

string

details

Additional error details

Example

{
  "code": "UNPROCESSABLE_ENTITY",
  "message": "Request could not be processed"
}

500

Internal server error

Media typeapplication/json

object

code

required

Error code

string

message

required

Human-readable error message

string

details

Additional error details

Example

{
  "code": "INTERNAL_ERROR",
  "message": "Internal server error"
}