GPU AI Service

List deployable Inventory items

get

Returns GPU offerings that are deployable right now based on live inventory, with pricing and availability classification.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Query parameters
regionstringOptional

Filter by region name (e.g., as-south-1)

typestring · enumOptional

Filter by item type (GPU only)

Default: GPUPossible values:
gpu_modelstringOptional

Filter GPU items by model (e.g., A40, A100, L40S)

gpu_vendorstringOptional

Filter by GPU vendor (case-insensitive, e.g., NVIDIA)

vram_minintegerOptional

Minimum VRAM (GiB). 0 or missing means no lower bound.

vram_maxintegerOptional

Maximum VRAM (GiB). 0 or missing means no upper bound.

pageinteger · min: 1Optional

Page number for pagination (starts from 1)

Default: 1
limitinteger · min: 1 · max: 100Optional

Number of items per page (max 100)

Default: 20
Responses
chevron-right
200

Inventory retrieved

application/json
Responseany
get
/api/v1beta1/inventory

No content

Get GPU config details by config_id

get
Authorizations
AuthorizationstringRequired

JWT token for authentication

Query parameters
config_idstringRequired

Unique config identifier from gpu_devices

Responses
chevron-right
200

GPU config retrieved

application/json
Responseany
get
/api/v1beta1/inventory/gpuconfig

No content

List GPU notification requests for the authenticated user

get
Authorizations
AuthorizationstringRequired

JWT token for authentication

Query parameters
statusstring · enumOptional

Filter requests by status

Possible values:
Responses
chevron-right
200

List of notification requests

application/json
ResponseGPUNotifyAvailabilityRequest[]
get
/api/v1beta1/inventory/gpu/notify-availability

Register for email notification when a GPU becomes available

post
Authorizations
AuthorizationstringRequired

JWT token for authentication

Body
region_idstringRequired

Region ID (e.g. as-south-1)

config_idstringRequired

GPU Config ID to watch for

Responses
post
/api/v1beta1/inventory/gpu/notify-availability

No content

List AI Runtimes

get

Retrieves a list of AI Runtimes for a specific project.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Path parameters
org_idstringRequired

The organization identifier.

project_idstringRequired

The project identifier.

Query parameters
pageinteger · min: 1Optional

Page number for pagination (starts from 1)

Default: 1
limitinteger · min: 1 · max: 100Optional

Number of items per page (max 100)

Default: 20
Responses
chevron-right
200

A paginated list of AI Runtimes.

application/json
Responseany
get
/api/v1beta1/orgs/{org_id}/projects/{project_id}/airuntimes

No content

Create an AI Runtime

post

Creates a new AI Runtime instance from a direct specification or a pre-defined template.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Path parameters
org_idstringRequired

The organization identifier.

project_idstringRequired

The project identifier.

Body
anyOptional
Responses
post
/api/v1beta1/orgs/{org_id}/projects/{project_id}/airuntimes

No content

Get AI Runtime by ID

get

Fetches the details of a specific AI Runtime by its unique identifier.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Path parameters
org_idstringRequired

The organization identifier.

project_idstringRequired

The project identifier.

airuntime_idstring · uuidRequired

The unique identifier of the AI Runtime.

Responses
chevron-right
200

Detailed information about the AI Runtime.

application/json
Responseany
get
/api/v1beta1/orgs/{org_id}/projects/{project_id}/airuntimes/{airuntime_id}

No content

Delete an AI Runtime

delete

Permanently deletes a specific AI Runtime by its unique ID.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Path parameters
org_idstringRequired

The organization identifier.

project_idstringRequired

The project identifier.

airuntime_idstring · uuidRequired

The unique identifier of the AI Runtime to delete.

Responses
delete
/api/v1beta1/orgs/{org_id}/projects/{project_id}/airuntimes/{airuntime_id}

No content

Get AI Runtime utilization metrics

get

Fetches real-time resource utilization metrics (CPU, RAM, GPU) for a specific AI Runtime.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Path parameters
org_idstringRequired

The organization identifier.

project_idstringRequired

The project identifier.

airuntime_idstring · uuidRequired

The unique identifier of the AI Runtime.

Responses
chevron-right
200

Resource utilization metrics for the AI Runtime.

application/json
Responseany
get
/api/v1beta1/orgs/{org_id}/projects/{project_id}/airuntimes/{airuntime_id}/metrics

No content

List AI Runtime Templates

get

Retrieves a list of available AI Runtime Templates.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Responses
chevron-right
200

A paginated list of AI Runtime Templates.

application/json
Responseany
get
/api/v1beta1/airuntime-templates

No content

Get an AI Runtime Template

get

Retrieves the details of a specific AI Runtime Template by its unique ID.

Authorizations
AuthorizationstringRequired

JWT token for authentication

Path parameters
template_idstringRequired

The unique identifier of the AI Runtime Template.

Pattern: ^tpl-[a-zA-Z0-9-]+$
Responses
chevron-right
200

Detailed information about the AI Runtime Template.

application/json
Responseany
get
/api/v1beta1/airuntime-templates/{template_id}

No content

Last updated