Skip to main content
The Together AI target adjusts the autoscaling range of one dedicated endpoint. Together AI continues choosing the live replica count within that range.

​
Configuration

Use the endpoint ID shown in Together AI, such as endpoint-abcd1234.

​
Authentication

Create an HTTP Auth Headers credential with the bearer token used for the Together AI API:

​
Observations and changes

Arklow reads the endpoint state and its configured autoscaling range. Together AI does not expose a live replica count through this connection, so Observed replicas remains empty and the target reports reduced observability. Scale changes update the endpoint’s minimum and maximum replicas together. Arklow pauses changes while the endpoint is STOPPED; start the endpoint in Together AI to resume scaling.

​
Add a Together AI target

1

Open scale targets

Go to Scale Targets, click New scale target, and select Together AI.
2

Identify the endpoint

Enter the Endpoint ID.
3

Select authentication

Choose the HTTP Auth Headers credential containing your Together AI bearer token.
4

Select capacity

Choose the destination or shared pool served by this endpoint.
5

Set the envelope

Enter the minimum and maximum replica counts available to the Together AI autoscaler.
6

Save

Create the target and confirm that Last observed advances and Health shows no credential or endpoint error.