Skip to main content
Inference providers enable you to connect external machine learning models and services to your Avala projects. Configure API endpoints, authentication, and test connectivity to ensure smooth integration.

​
List inference providers

Retrieve a paginated list of inference providers. 
const providers = await avala.inferenceProviders.list({
  limit: 50,
  cursor: 'next-cursor'
});

​
Parameters

​
options.limit
number
Maximum number of providers to return per page
​
options.cursor
string
Cursor for pagination

​
Returns

Returns a cursor-paginated list of inference provider objects.
​
items
InferenceProvider[]
Array of inference provider objects
​
nextCursor
string | null
Cursor for the next page of results
​
previousCursor
string | null
Cursor for the previous page of results
​
hasMore
boolean
Whether more results are available

​
Create an inference provider

Create a new inference provider. 
const provider = await avala.inferenceProviders.create({
  name: 'YOLOv8 Detection',
  description: 'Object detection using YOLOv8',
  providerType: 'http',
  config: {
    endpoint: 'https://api.example.com/predict',
    apiKey: 'your-api-key',
    timeout: 30000
  },
  project: 'project-uid'
});

​
Parameters

​
options.name
string
required
Name of the inference provider
​
options.description
string
Description of the provider’s purpose
​
options.providerType
string
required
Type of provider (e.g., ‘http’, ‘grpc’, ‘aws’, ‘gcp’)
​
options.config
Record<string, unknown>
required
Configuration object containing provider-specific settings like endpoint URL, API keys, etc.
​
options.project
string
Project UID to scope the provider to

​
Returns

Returns the created inference provider object.
​
uid
string
Unique identifier for the provider
​
name
string
Name of the provider
​
description
string | null
Description of the provider
​
providerType
string | null
Type of provider
​
config
object | null
Configuration settings
​
isActive
boolean
Whether the provider is active
​
project
string | null
Project UID the provider is scoped to
​
lastTestAt
string | null
ISO 8601 timestamp of last connectivity test
​
lastTestSuccess
boolean | null
Whether the last test was successful
​
createdAt
string | null
ISO 8601 timestamp of creation
​
updatedAt
string | null
ISO 8601 timestamp of last update

​
Get an inference provider

Retrieve a specific inference provider by UID. 
const provider = await avala.inferenceProviders.get('provider-uid');

​
Parameters

​
uid
string
required
The unique identifier of the inference provider

​
Returns

Returns the inference provider object.

​
Update an inference provider

Update an existing inference provider. 
const provider = await avala.inferenceProviders.update('provider-uid', {
  name: 'Updated Provider Name',
  isActive: false,
  config: {
    endpoint: 'https://api.example.com/v2/predict',
    apiKey: 'new-api-key'
  }
});

​
Parameters

​
uid
string
required
The unique identifier of the provider to update
​
options.name
string
New name for the provider
​
options.description
string
New description
​
options.providerType
string
New provider type
​
options.config
Record<string, unknown>
Updated configuration settings
​
options.isActive
boolean
Whether the provider should be active
​
options.project
string
New project UID to scope the provider to

​
Returns

Returns the updated inference provider object.

​
Delete an inference provider

Delete an inference provider. 
await avala.inferenceProviders.delete('provider-uid');

​
Parameters

​
uid
string
required
The unique identifier of the provider to delete

​
Returns

Returns void on success.

​
Test an inference provider

Test connectivity and configuration of an inference provider. 
const result = await avala.inferenceProviders.test('provider-uid');

​
Parameters

​
uid
string
required
The unique identifier of the provider to test

​
Returns

​
success
boolean
Whether the test was successful
​
message
string
Description of the test result
​
testedAt
string
ISO 8601 timestamp of when the test was performed

Build docs developers (and LLMs) love

Get started for freeTalk to us