Skip to main content
Applies to:
  • Plan -
  • Deployment -

​
Summary

Loop’s models are served either by Braintrust’s built-in models or by one of your organization’s own AI providers. In the model picker, select Braintrust to use built-in models, or select one of your organization’s own providers. When neither is available, the picker shows Add provider or a No org-level provider warning. Configure a compatible organization-level provider, or re-enable built-in models.

​
What is happening

Loop resolves a model in one of two ways:
  • Built-in models. Braintrust serves the model and usage draws down your organization’s monthly model credits. This is the default, and it requires no provider setup.
  • Your own provider. The model runs through one of your organization’s OpenAI-compatible AI providers, and inference is billed to that provider.
Three constraints cause most reports of a missing provider or model:
  • Loop uses organization-level AI providers only. A provider configured at the project level serves playgrounds, prompts, and the gateway, but not Loop.
  • Loop offers only its supported models. Other models from your providers are not selectable in Loop, so adding a custom model to a provider does not make it available in a Loop thread.
  • On the Starter plan, built-in models require at least one organization owner with a work email address, or a payment method on file. An organization that doesn’t qualify cannot select Braintrust as a provider, even with Allow built-in models on. See Requirements.
Compatible provider types are OpenAI, Azure, Bedrock, OpenRouter, and Azure AI Gateway. Your configured provider must serve the Loop model you select.

​
Resolve it

​
Check whether built-in models are allowed

  1. Go to Settings > AI providers.
  2. Under Built-in models, check Allow built-in models.
If it is on, Braintrust appears as a provider option in Loop, unless your organization is on Starter and doesn’t meet the work email or payment method requirement. Self-hosted organizations have built-in models off by default. See Manage built-in models. If it is off by choice, or your model credits are exhausted, continue below.

​
Configure a compatible organization-level provider

  1. On the same page, add an Organization provider of a compatible type.
  2. Save the provider, then reload the Loop page.
You can also configure a custom endpoint directly from Loop. Click Add provider in the model picker, then click New under Custom providers. Choose a compatible provider type and configure an endpoint that serves the Loop model you want to use. Only members of the Owners permission group, or a custom permission group with the Manage settings organization permission, can add organization providers or change the built-in models setting.

​
How to confirm it worked

  • Open Loop and open the model picker.
  • The picker shows a model name rather than Add provider or No org-level provider.
  • The Provider setting lists Braintrust, your configured providers, or both, depending on your organization settings. Select a provider and run a short prompt to confirm it can serve the model.

​
Notes

  • A project-level provider does not resolve this. Reconfigure it at the organization level, or add a second organization-level provider.
  • Turning Allow built-in models off pauses your topic automations. Re-enabling does not resume them automatically. See Manage built-in models.
  • Once your model credits are exhausted, Braintrust is unavailable as a provider, so an organization-level provider is required to keep using Loop.