AI Providers

Configure LLM providers and custom models for evaluations, optimization, and other platform features.

About

AI Providers is where you connect LLM services to Future AGI. The platform uses these providers for evaluations, prototype runs, optimization, and other features that need to call a language model. You can add built-in providers (like OpenAI, Anthropic, AWS Bedrock), cloud providers (like Azure OpenAI), or configure custom model endpoints.

Access: Owner and Admin at the organization level. Workspace admins and members can view and use configured providers.

Built-in Providers

These providers are pre-configured. You just need to add your API key.

Common providers include OpenAI, Anthropic, Google (Gemini/Vertex AI), AWS Bedrock, Azure OpenAI, Mistral, Cohere, and others.

How to Add a Provider

  1. Go to Settings > AI Providers
  2. Click on the provider you want to add (or click Create custom model for a custom endpoint)
  3. Enter your API key and any required configuration (region, endpoint, etc.)
  4. Click Save

Custom Models

For self-hosted models or custom API endpoints:

  1. Click Create custom model
  2. Enter the following details:
    • Model name - display name for the platform
    • API base URL - the endpoint Future AGI will call
    • Any custom headers or authentication parameters
  3. Click Save

Tip

The custom model then appears alongside built-in providers when selecting a model for evaluations, optimization, or other features.

Filtering Providers

Use the filter buttons to narrow the view:

FilterShows
All ProvidersEverything configured
Default Model ProvidersStandard LLM providers (OpenAI, Anthropic, etc.)
Cloud ProvidersCloud-hosted providers (AWS Bedrock, Azure OpenAI, etc.)
Custom ModelsYour custom model endpoints

Workspace-Level Providers

Each workspace can have its own AI provider configuration. Go to Settings > Workspace > [workspace name] > AI Providers to configure workspace-specific providers. Workspace providers override organization-level providers for that workspace.

Next Steps

Was this page helpful?

Questions & Discussion