Skip to content

Saved AIs

The AI Readiness routes for saved ais. The other areas are listed on the AI Readiness reference.

Every route below follows the conventions in Work with the BI Pixie API. You can also download the OpenAPI document.

List the account's saved AIs #

GET /v1/aiReadiness/aiProviders

Returns the AIs the account has saved, as the portal's AI setup page lists them: the one BI Pixie writes optimizations with, the ones a benchmark can test as a generic AI agent, and the saved credentials they use. No answer carries a key. The saved AIs are an account admin's to see and change; a member is answered 403 AdminRoleRequired.

Scope
BIPixie.Read.All
Plan
Every plan
Counts against
Calls that read, 300 a minute

Responses

200. The saved AIs.

Headers of the 200 response to List the account's saved AIs
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 200 response to List the account's saved AIs, SavedAiList
Field Type Always present Description
value array of SavedAi Yes
providerId string Yes The id a benchmark run's GenericAgent takes.
name string or null No The name the account gave it.
kind string Yes The service. One of openai, anthropic, azure_openai, fabric_data_agent.
model string or null No
deployment string or null No The Azure AI Foundry deployment, for azure_openai.
credentialId string or null No The saved credential it uses. None for a data agent, which needs no key.
workspaceId string or null No The Fabric workspace of a data agent.
dataAgentId string or null No The data agent's Fabric item id.
dataAgentName string or null No
isWriter boolean Yes Whether this is the AI BI Pixie writes optimizations with.
isTested boolean Yes Whether a benchmark can test it as a generic AI agent.
reachable boolean Yes Whether BI Pixie could reach it the last time it tried.
addedAtUtc string or null (date-time) No
credentials array of SavedAiCredential Yes
credentialId string Yes
kind string Yes One of openai, anthropic, azure_openai.
authMode string Yes How BI Pixie signs in to the service. entra is for Azure AI Foundry only. One of api_key, entra.
endpoint string or null No
tenantId string or null No For entra only.
clientId string or null No For entra only.
usedBy integer Yes How many saved AIs use it.
addedAtUtc string or null (date-time) No
maxProviders integer Yes The most AIs an account can save.

Errors

Every refusal is an ErrorResponse with a stable errorCode. See Errors for what each HTTP status means and every code the API can return.

This is an example of a refusal. Every refusal has the same fields.

Example refusal 
HTTP/1.1 404 Not Found
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "errorCode": "ItemNotFound",
  "message": "No item with this id is visible to the signed-in person in this workspace, or it is not of the type the route takes.",
  "isRetriable": false,
  "moreDetails": [],
  "relatedResource": {
    "resourceId": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa",
    "resourceType": "SemanticModel"
  },
  "parameters": [
    {
      "name": "itemId",
      "value": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa"
    }
  ],
  "requestId": "cbf21b6d-b661-4b52-88cb-d79712e504fc"
}

Example

Request 
curl -i -X GET \
  -H "Authorization: Bearer $TOKEN" \
  "https://api.bipixie.com/v1/aiReadiness/aiProviders"

An example of what the call answers. The values are fictional.

Response, 200 
HTTP/1.1 200 OK
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "value": [
    {
      "providerId": "b2df6be3-7769-4222-90ea-9fd98f4592e0",
      "name": "Contoso Sales",
      "kind": "openai",
      "model": "Contoso Sales",
      "deployment": "Contoso Sales",
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "workspaceId": "e17c1b59-7a2e-491e-acdd-11434854bbb0",
      "dataAgentId": "05dd87d1-624f-4ffa-ab88-ee231a75a904",
      "dataAgentName": "Contoso Sales Agent",
      "isWriter": true,
      "isTested": true,
      "reachable": true,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "credentials": [
    {
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "kind": "openai",
      "authMode": "api_key",
      "endpoint": "Contoso Sales",
      "tenantId": "4b0e5e52-dfc8-4417-8d4a-94a8e7d7f08d",
      "clientId": "ebc40edd-1874-4e14-a1e5-e21b45f6286a",
      "usedBy": 12,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "maxProviders": 3
}

Save an AI #

POST /v1/aiReadiness/aiProviders

Saves an AI, with a new key in the body or by naming a saved credential with credentialId, as the setup page does for a second Azure AI Foundry deployment on the same resource. With role optimizations it becomes the AI BI Pixie writes with, and the AI it replaces leaves the list unless it is also tested or keepReplaced is true; with role test a benchmark can test it as a generic AI agent. The first AI saved is also the one BI Pixie writes with. An AI that is already saved is given the other role rather than saved twice. An account holds at most ten saved AIs. A key is taken in the body only: a request whose address carries one is refused 400 InvalidRequest, and no answer returns it. Answers the whole list, because adding one can change another.

Scope
BIPixie.ReadWrite.All
Plan
Every plan
Counts against
Calls that start work, 30 a minute

Request body

JSON, AddSavedAiRequest. The body is required.

Request body for Save an AI
Field Type Required Description
kind string Yes One of openai, anthropic, azure_openai, fabric_data_agent.
name string No
role string No optimizations makes it the AI BI Pixie writes with; test lets a benchmark test it. When left out, a data agent is saved as the writer and any other AI as one to test, except that the first AI saved is also the writer. One of optimizations, test.
keepReplaced boolean No With role optimizations, keep the AI it replaces in the list as one to test.
model string No
deployment string No The Azure AI Foundry deployment, for azure_openai.
credentialId string No A saved credential of the same service, instead of a new key.
authMode string No One of api_key, entra.
endpoint string No
apiKey string No The key. Encrypted when saved and never returned.
clientSecret string No The client secret, for entra. Encrypted when saved and never returned.
tenantId string No
clientId string No
workspaceId string No The Fabric workspace of a data agent.
dataAgentId string No
dataAgentName string No

Responses

201. Saved. The body is the whole list as it now stands.

Headers of the 201 response to Save an AI
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 201 response to Save an AI, SavedAiList
Field Type Always present Description
value array of SavedAi Yes
providerId string Yes The id a benchmark run's GenericAgent takes.
name string or null No The name the account gave it.
kind string Yes The service. One of openai, anthropic, azure_openai, fabric_data_agent.
model string or null No
deployment string or null No The Azure AI Foundry deployment, for azure_openai.
credentialId string or null No The saved credential it uses. None for a data agent, which needs no key.
workspaceId string or null No The Fabric workspace of a data agent.
dataAgentId string or null No The data agent's Fabric item id.
dataAgentName string or null No
isWriter boolean Yes Whether this is the AI BI Pixie writes optimizations with.
isTested boolean Yes Whether a benchmark can test it as a generic AI agent.
reachable boolean Yes Whether BI Pixie could reach it the last time it tried.
addedAtUtc string or null (date-time) No
credentials array of SavedAiCredential Yes
credentialId string Yes
kind string Yes One of openai, anthropic, azure_openai.
authMode string Yes How BI Pixie signs in to the service. entra is for Azure AI Foundry only. One of api_key, entra.
endpoint string or null No
tenantId string or null No For entra only.
clientId string or null No For entra only.
usedBy integer Yes How many saved AIs use it.
addedAtUtc string or null (date-time) No
maxProviders integer Yes The most AIs an account can save.

400. The request is not well formed, and nothing was started, changed or counted. errorCode is InvalidRequest, or InvalidPreview for a preview that is not available. A body member the operation does not document is refused this way, and the message names it and the documented member closest to it.

Headers of the 400 response to Save an AI
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 400 response to Save an AI, ErrorResponse
Field Type Always present Description
errorCode string Yes Each code keeps its meaning once published. New codes can be added over time, and a code is never reused. One of InvalidRequest, InvalidPreview, PayloadTooLarge, BenchmarkQuestionsMissing, OperationNotSupportedForItem, Unauthorized, TokenExpired, InsufficientScopes, AppOnlyTokenNotSupported, UserNotRegistered, AccountNotActive, AdminRoleRequired, FeatureNotInPlan, FeatureTurnedOff, LicenseNotYours, TrackedItemLimitReached, LicenseTrackedItemLimitReached, FabricConsentRequired, FabricAccessDenied, WorkspaceNotFound, ItemNotFound, AssessmentNotFound, RunNotFound, AbTestNotFound, OperationNotFound, ScheduleNotFound, AiProviderNotFound, AiCredentialNotFound, OperationInProgress, OperationHasNoResult, ResultNotReady, ScheduleConflict, AiProviderConflict, ItemChanged, CapacityPaused, OperationExpired, RegionNotReady, CapacityRequired, CapacityUnpaid, StoreNotConnected, RequestBlocked, AssessmentDailyLimitReached, BenchmarkRunsExhausted, OptimizationApplyDailyLimitReached, FirstLookLimitReached, BenchmarkBuildLimitReached, FabricUnavailable, StoreUnavailable, ThrottleStoreUnavailable, UnknownError.
message string Yes
isRetriable boolean No
moreDetails array of object No
relatedResource object or null No
resourceId string No
resourceType string No
parameters array of object No
name string Yes
value string Yes
requestId string (uuid) Yes

Errors

Every refusal is an ErrorResponse with a stable errorCode. See Errors for what each HTTP status means and every code the API can return.

This is an example of a refusal. Every refusal has the same fields.

Example refusal 
HTTP/1.1 404 Not Found
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "errorCode": "ItemNotFound",
  "message": "No item with this id is visible to the signed-in person in this workspace, or it is not of the type the route takes.",
  "isRetriable": false,
  "moreDetails": [],
  "relatedResource": {
    "resourceId": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa",
    "resourceType": "SemanticModel"
  },
  "parameters": [
    {
      "name": "itemId",
      "value": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa"
    }
  ],
  "requestId": "cbf21b6d-b661-4b52-88cb-d79712e504fc"
}

Example

Request 
curl -i -X POST \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
  "kind": "openai"
}' \
  "https://api.bipixie.com/v1/aiReadiness/aiProviders"

An example of what the call answers. The values are fictional.

Response, 201 
HTTP/1.1 201 Created
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "value": [
    {
      "providerId": "b2df6be3-7769-4222-90ea-9fd98f4592e0",
      "name": "Contoso Sales",
      "kind": "openai",
      "model": "Contoso Sales",
      "deployment": "Contoso Sales",
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "workspaceId": "e17c1b59-7a2e-491e-acdd-11434854bbb0",
      "dataAgentId": "05dd87d1-624f-4ffa-ab88-ee231a75a904",
      "dataAgentName": "Contoso Sales Agent",
      "isWriter": true,
      "isTested": true,
      "reachable": true,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "credentials": [
    {
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "kind": "openai",
      "authMode": "api_key",
      "endpoint": "Contoso Sales",
      "tenantId": "4b0e5e52-dfc8-4417-8d4a-94a8e7d7f08d",
      "clientId": "ebc40edd-1874-4e14-a1e5-e21b45f6286a",
      "usedBy": 12,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "maxProviders": 3
}

Rename a saved AI, or make it the one BI Pixie writes with #

PATCH /v1/aiReadiness/aiProviders/{providerId}

Renames the saved AI with name, makes it the AI BI Pixie writes optimizations with when isWriter is true, or both. The AI it replaces leaves the list unless it is also tested. The model is not changed here: a different model is a different saved AI. Answers the whole list, because making one the writer can change another.

Scope
BIPixie.ReadWrite.All
Plan
Every plan
Counts against
Calls that start work, 30 a minute

Parameters

Parameters for Rename a saved AI, or make it the one BI Pixie writes with
Name In Type Required Description
providerId path string Yes The saved AI's id, as the saved AIs list gives it. A benchmark run's GenericAgent takes the same id.

Request body

JSON, UpdateSavedAiRequest. The body is required.

Request body for Rename a saved AI, or make it the one BI Pixie writes with
Field Type Required Description
name string No
isWriter boolean No True makes it the AI BI Pixie writes with. False changes nothing.

Responses

200. Saved. The body is the whole list as it now stands.

Headers of the 200 response to Rename a saved AI, or make it the one BI Pixie writes with
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 200 response to Rename a saved AI, or make it the one BI Pixie writes with, SavedAiList
Field Type Always present Description
value array of SavedAi Yes
providerId string Yes The id a benchmark run's GenericAgent takes.
name string or null No The name the account gave it.
kind string Yes The service. One of openai, anthropic, azure_openai, fabric_data_agent.
model string or null No
deployment string or null No The Azure AI Foundry deployment, for azure_openai.
credentialId string or null No The saved credential it uses. None for a data agent, which needs no key.
workspaceId string or null No The Fabric workspace of a data agent.
dataAgentId string or null No The data agent's Fabric item id.
dataAgentName string or null No
isWriter boolean Yes Whether this is the AI BI Pixie writes optimizations with.
isTested boolean Yes Whether a benchmark can test it as a generic AI agent.
reachable boolean Yes Whether BI Pixie could reach it the last time it tried.
addedAtUtc string or null (date-time) No
credentials array of SavedAiCredential Yes
credentialId string Yes
kind string Yes One of openai, anthropic, azure_openai.
authMode string Yes How BI Pixie signs in to the service. entra is for Azure AI Foundry only. One of api_key, entra.
endpoint string or null No
tenantId string or null No For entra only.
clientId string or null No For entra only.
usedBy integer Yes How many saved AIs use it.
addedAtUtc string or null (date-time) No
maxProviders integer Yes The most AIs an account can save.

400. The request is not well formed, and nothing was started, changed or counted. errorCode is InvalidRequest, or InvalidPreview for a preview that is not available. A body member the operation does not document is refused this way, and the message names it and the documented member closest to it.

Headers of the 400 response to Rename a saved AI, or make it the one BI Pixie writes with
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 400 response to Rename a saved AI, or make it the one BI Pixie writes with, ErrorResponse
Field Type Always present Description
errorCode string Yes Each code keeps its meaning once published. New codes can be added over time, and a code is never reused. One of InvalidRequest, InvalidPreview, PayloadTooLarge, BenchmarkQuestionsMissing, OperationNotSupportedForItem, Unauthorized, TokenExpired, InsufficientScopes, AppOnlyTokenNotSupported, UserNotRegistered, AccountNotActive, AdminRoleRequired, FeatureNotInPlan, FeatureTurnedOff, LicenseNotYours, TrackedItemLimitReached, LicenseTrackedItemLimitReached, FabricConsentRequired, FabricAccessDenied, WorkspaceNotFound, ItemNotFound, AssessmentNotFound, RunNotFound, AbTestNotFound, OperationNotFound, ScheduleNotFound, AiProviderNotFound, AiCredentialNotFound, OperationInProgress, OperationHasNoResult, ResultNotReady, ScheduleConflict, AiProviderConflict, ItemChanged, CapacityPaused, OperationExpired, RegionNotReady, CapacityRequired, CapacityUnpaid, StoreNotConnected, RequestBlocked, AssessmentDailyLimitReached, BenchmarkRunsExhausted, OptimizationApplyDailyLimitReached, FirstLookLimitReached, BenchmarkBuildLimitReached, FabricUnavailable, StoreUnavailable, ThrottleStoreUnavailable, UnknownError.
message string Yes
isRetriable boolean No
moreDetails array of object No
relatedResource object or null No
resourceId string No
resourceType string No
parameters array of object No
name string Yes
value string Yes
requestId string (uuid) Yes

Errors

Every refusal is an ErrorResponse with a stable errorCode. See Errors for what each HTTP status means and every code the API can return.

This is an example of a refusal. Every refusal has the same fields.

Example refusal 
HTTP/1.1 404 Not Found
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "errorCode": "ItemNotFound",
  "message": "No item with this id is visible to the signed-in person in this workspace, or it is not of the type the route takes.",
  "isRetriable": false,
  "moreDetails": [],
  "relatedResource": {
    "resourceId": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa",
    "resourceType": "SemanticModel"
  },
  "parameters": [
    {
      "name": "itemId",
      "value": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa"
    }
  ],
  "requestId": "cbf21b6d-b661-4b52-88cb-d79712e504fc"
}

Example

Request 
curl -i -X PATCH \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
  "name": "<string>",
  "isWriter": false
}' \
  "https://api.bipixie.com/v1/aiReadiness/aiProviders/$PROVIDER_ID"

An example of what the call answers. The values are fictional.

Response, 200 
HTTP/1.1 200 OK
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "value": [
    {
      "providerId": "b2df6be3-7769-4222-90ea-9fd98f4592e0",
      "name": "Contoso Sales",
      "kind": "openai",
      "model": "Contoso Sales",
      "deployment": "Contoso Sales",
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "workspaceId": "e17c1b59-7a2e-491e-acdd-11434854bbb0",
      "dataAgentId": "05dd87d1-624f-4ffa-ab88-ee231a75a904",
      "dataAgentName": "Contoso Sales Agent",
      "isWriter": true,
      "isTested": true,
      "reachable": true,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "credentials": [
    {
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "kind": "openai",
      "authMode": "api_key",
      "endpoint": "Contoso Sales",
      "tenantId": "4b0e5e52-dfc8-4417-8d4a-94a8e7d7f08d",
      "clientId": "ebc40edd-1874-4e14-a1e5-e21b45f6286a",
      "usedBy": 12,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "maxProviders": 3
}

Remove a saved AI #

DELETE /v1/aiReadiness/aiProviders/{providerId}

Removes the saved AI. With role test it is only taken out of the AIs a benchmark can test, and stays the AI BI Pixie writes with if it is that; with role optimizations it is removed only if it is the AI BI Pixie writes with. The AI BI Pixie writes with is changed rather than removed while another AI is saved (409 AiProviderConflict, reason IsTheWriter); the last saved AI can be removed, which leaves the account with none. A credential no saved AI uses any more is removed with it. Answers the whole list.

Scope
BIPixie.ReadWrite.All
Plan
Every plan
Counts against
Calls that start work, 30 a minute

Parameters

Parameters for Remove a saved AI
Name In Type Required Description
providerId path string Yes The saved AI's id, as the saved AIs list gives it. A benchmark run's GenericAgent takes the same id.
role query string No Take the AI out of this one section only. Both, when left out. One of optimizations, test.

Responses

200. Removed. The body is the whole list as it now stands.

Headers of the 200 response to Remove a saved AI
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 200 response to Remove a saved AI, SavedAiList
Field Type Always present Description
value array of SavedAi Yes
providerId string Yes The id a benchmark run's GenericAgent takes.
name string or null No The name the account gave it.
kind string Yes The service. One of openai, anthropic, azure_openai, fabric_data_agent.
model string or null No
deployment string or null No The Azure AI Foundry deployment, for azure_openai.
credentialId string or null No The saved credential it uses. None for a data agent, which needs no key.
workspaceId string or null No The Fabric workspace of a data agent.
dataAgentId string or null No The data agent's Fabric item id.
dataAgentName string or null No
isWriter boolean Yes Whether this is the AI BI Pixie writes optimizations with.
isTested boolean Yes Whether a benchmark can test it as a generic AI agent.
reachable boolean Yes Whether BI Pixie could reach it the last time it tried.
addedAtUtc string or null (date-time) No
credentials array of SavedAiCredential Yes
credentialId string Yes
kind string Yes One of openai, anthropic, azure_openai.
authMode string Yes How BI Pixie signs in to the service. entra is for Azure AI Foundry only. One of api_key, entra.
endpoint string or null No
tenantId string or null No For entra only.
clientId string or null No For entra only.
usedBy integer Yes How many saved AIs use it.
addedAtUtc string or null (date-time) No
maxProviders integer Yes The most AIs an account can save.

400. The request is not well formed, and nothing was started, changed or counted. errorCode is InvalidRequest, or InvalidPreview for a preview that is not available. A body member the operation does not document is refused this way, and the message names it and the documented member closest to it.

Headers of the 400 response to Remove a saved AI
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 400 response to Remove a saved AI, ErrorResponse
Field Type Always present Description
errorCode string Yes Each code keeps its meaning once published. New codes can be added over time, and a code is never reused. One of InvalidRequest, InvalidPreview, PayloadTooLarge, BenchmarkQuestionsMissing, OperationNotSupportedForItem, Unauthorized, TokenExpired, InsufficientScopes, AppOnlyTokenNotSupported, UserNotRegistered, AccountNotActive, AdminRoleRequired, FeatureNotInPlan, FeatureTurnedOff, LicenseNotYours, TrackedItemLimitReached, LicenseTrackedItemLimitReached, FabricConsentRequired, FabricAccessDenied, WorkspaceNotFound, ItemNotFound, AssessmentNotFound, RunNotFound, AbTestNotFound, OperationNotFound, ScheduleNotFound, AiProviderNotFound, AiCredentialNotFound, OperationInProgress, OperationHasNoResult, ResultNotReady, ScheduleConflict, AiProviderConflict, ItemChanged, CapacityPaused, OperationExpired, RegionNotReady, CapacityRequired, CapacityUnpaid, StoreNotConnected, RequestBlocked, AssessmentDailyLimitReached, BenchmarkRunsExhausted, OptimizationApplyDailyLimitReached, FirstLookLimitReached, BenchmarkBuildLimitReached, FabricUnavailable, StoreUnavailable, ThrottleStoreUnavailable, UnknownError.
message string Yes
isRetriable boolean No
moreDetails array of object No
relatedResource object or null No
resourceId string No
resourceType string No
parameters array of object No
name string Yes
value string Yes
requestId string (uuid) Yes

Errors

Every refusal is an ErrorResponse with a stable errorCode. See Errors for what each HTTP status means and every code the API can return.

This is an example of a refusal. Every refusal has the same fields.

Example refusal 
HTTP/1.1 404 Not Found
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "errorCode": "ItemNotFound",
  "message": "No item with this id is visible to the signed-in person in this workspace, or it is not of the type the route takes.",
  "isRetriable": false,
  "moreDetails": [],
  "relatedResource": {
    "resourceId": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa",
    "resourceType": "SemanticModel"
  },
  "parameters": [
    {
      "name": "itemId",
      "value": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa"
    }
  ],
  "requestId": "cbf21b6d-b661-4b52-88cb-d79712e504fc"
}

Example

Request 
curl -i -X DELETE \
  -H "Authorization: Bearer $TOKEN" \
  "https://api.bipixie.com/v1/aiReadiness/aiProviders/$PROVIDER_ID"

An example of what the call answers. The values are fictional.

Response, 200 
HTTP/1.1 200 OK
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "value": [
    {
      "providerId": "b2df6be3-7769-4222-90ea-9fd98f4592e0",
      "name": "Contoso Sales",
      "kind": "openai",
      "model": "Contoso Sales",
      "deployment": "Contoso Sales",
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "workspaceId": "e17c1b59-7a2e-491e-acdd-11434854bbb0",
      "dataAgentId": "05dd87d1-624f-4ffa-ab88-ee231a75a904",
      "dataAgentName": "Contoso Sales Agent",
      "isWriter": true,
      "isTested": true,
      "reachable": true,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "credentials": [
    {
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "kind": "openai",
      "authMode": "api_key",
      "endpoint": "Contoso Sales",
      "tenantId": "4b0e5e52-dfc8-4417-8d4a-94a8e7d7f08d",
      "clientId": "ebc40edd-1874-4e14-a1e5-e21b45f6286a",
      "usedBy": 12,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "maxProviders": 3
}

Replace a saved credential's key #

PUT /v1/aiReadiness/aiProviderCredentials/{credentialId}

Replaces the key of a saved credential, which changes every saved AI that uses it, and clears what BI Pixie last recorded about reaching them. The service, the sign-in method and the endpoint stay: a different resource is a different credential. A credential that signs in with Microsoft Entra takes a new clientSecret, and may take a new tenantId or clientId; any other takes a new apiKey. The key is taken in the body only, and no answer returns it. Answers the whole list.

Scope
BIPixie.ReadWrite.All
Plan
Every plan
Counts against
Calls that start work, 30 a minute

Parameters

Parameters for Replace a saved credential's key
Name In Type Required Description
credentialId path string Yes The saved credential's id, as the saved AIs list gives it.

Request body

JSON, ReplaceSavedAiKeyRequest. The body is required.

Request body for Replace a saved credential's key
Field Type Required Description
apiKey string No
clientSecret string No
tenantId string No
clientId string No

Responses

200. Replaced. The body is the whole list as it now stands.

Headers of the 200 response to Replace a saved credential's key
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 200 response to Replace a saved credential's key, SavedAiList
Field Type Always present Description
value array of SavedAi Yes
providerId string Yes The id a benchmark run's GenericAgent takes.
name string or null No The name the account gave it.
kind string Yes The service. One of openai, anthropic, azure_openai, fabric_data_agent.
model string or null No
deployment string or null No The Azure AI Foundry deployment, for azure_openai.
credentialId string or null No The saved credential it uses. None for a data agent, which needs no key.
workspaceId string or null No The Fabric workspace of a data agent.
dataAgentId string or null No The data agent's Fabric item id.
dataAgentName string or null No
isWriter boolean Yes Whether this is the AI BI Pixie writes optimizations with.
isTested boolean Yes Whether a benchmark can test it as a generic AI agent.
reachable boolean Yes Whether BI Pixie could reach it the last time it tried.
addedAtUtc string or null (date-time) No
credentials array of SavedAiCredential Yes
credentialId string Yes
kind string Yes One of openai, anthropic, azure_openai.
authMode string Yes How BI Pixie signs in to the service. entra is for Azure AI Foundry only. One of api_key, entra.
endpoint string or null No
tenantId string or null No For entra only.
clientId string or null No For entra only.
usedBy integer Yes How many saved AIs use it.
addedAtUtc string or null (date-time) No
maxProviders integer Yes The most AIs an account can save.

400. The request is not well formed, and nothing was started, changed or counted. errorCode is InvalidRequest, or InvalidPreview for a preview that is not available. A body member the operation does not document is refused this way, and the message names it and the documented member closest to it.

Headers of the 400 response to Replace a saved credential's key
Header Description
x-ms-request-id The id of this request. Quote it to support.
The 400 response to Replace a saved credential's key, ErrorResponse
Field Type Always present Description
errorCode string Yes Each code keeps its meaning once published. New codes can be added over time, and a code is never reused. One of InvalidRequest, InvalidPreview, PayloadTooLarge, BenchmarkQuestionsMissing, OperationNotSupportedForItem, Unauthorized, TokenExpired, InsufficientScopes, AppOnlyTokenNotSupported, UserNotRegistered, AccountNotActive, AdminRoleRequired, FeatureNotInPlan, FeatureTurnedOff, LicenseNotYours, TrackedItemLimitReached, LicenseTrackedItemLimitReached, FabricConsentRequired, FabricAccessDenied, WorkspaceNotFound, ItemNotFound, AssessmentNotFound, RunNotFound, AbTestNotFound, OperationNotFound, ScheduleNotFound, AiProviderNotFound, AiCredentialNotFound, OperationInProgress, OperationHasNoResult, ResultNotReady, ScheduleConflict, AiProviderConflict, ItemChanged, CapacityPaused, OperationExpired, RegionNotReady, CapacityRequired, CapacityUnpaid, StoreNotConnected, RequestBlocked, AssessmentDailyLimitReached, BenchmarkRunsExhausted, OptimizationApplyDailyLimitReached, FirstLookLimitReached, BenchmarkBuildLimitReached, FabricUnavailable, StoreUnavailable, ThrottleStoreUnavailable, UnknownError.
message string Yes
isRetriable boolean No
moreDetails array of object No
relatedResource object or null No
resourceId string No
resourceType string No
parameters array of object No
name string Yes
value string Yes
requestId string (uuid) Yes

Errors

Every refusal is an ErrorResponse with a stable errorCode. See Errors for what each HTTP status means and every code the API can return.

This is an example of a refusal. Every refusal has the same fields.

Example refusal 
HTTP/1.1 404 Not Found
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "errorCode": "ItemNotFound",
  "message": "No item with this id is visible to the signed-in person in this workspace, or it is not of the type the route takes.",
  "isRetriable": false,
  "moreDetails": [],
  "relatedResource": {
    "resourceId": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa",
    "resourceType": "SemanticModel"
  },
  "parameters": [
    {
      "name": "itemId",
      "value": "d972b3f5-a2fa-47ec-94ff-0403a4d159fa"
    }
  ],
  "requestId": "cbf21b6d-b661-4b52-88cb-d79712e504fc"
}

Example

Request 
curl -i -X PUT \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
  "apiKey": "<string>",
  "clientSecret": "<string>",
  "tenantId": "<string>",
  "clientId": "<string>"
}' \
  "https://api.bipixie.com/v1/aiReadiness/aiProviderCredentials/$CREDENTIAL_ID"

An example of what the call answers. The values are fictional.

Response, 200 
HTTP/1.1 200 OK
Content-Type: application/json
x-ms-request-id: cbf21b6d-b661-4b52-88cb-d79712e504fc

{
  "value": [
    {
      "providerId": "b2df6be3-7769-4222-90ea-9fd98f4592e0",
      "name": "Contoso Sales",
      "kind": "openai",
      "model": "Contoso Sales",
      "deployment": "Contoso Sales",
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "workspaceId": "e17c1b59-7a2e-491e-acdd-11434854bbb0",
      "dataAgentId": "05dd87d1-624f-4ffa-ab88-ee231a75a904",
      "dataAgentName": "Contoso Sales Agent",
      "isWriter": true,
      "isTested": true,
      "reachable": true,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "credentials": [
    {
      "credentialId": "5aaabc9b-2de4-43a2-91b4-fb5db999ba8c",
      "kind": "openai",
      "authMode": "api_key",
      "endpoint": "Contoso Sales",
      "tenantId": "4b0e5e52-dfc8-4417-8d4a-94a8e7d7f08d",
      "clientId": "ebc40edd-1874-4e14-a1e5-e21b45f6286a",
      "usedBy": 12,
      "addedAtUtc": "2026-09-14T09:30:05Z"
    }
  ],
  "maxProviders": 3
}

What's next