Providers

providers are how amberflo securely connects to the llm providers you use they form the foundation for unified access across providers and enable you to control how your applications interact with those providers without exposing provider specific credentials once a provider is set up in amberflo, it can be reused to grant fine grained access to models through workloads and access keys applications never receive direct provider api keys all access is mediated through the amberflo gateway why providers matter providers allow you to centralize access to multiple llm providers, like openai, aws bedrock, etc avoid embedding provider api keys in application code rotate or revoke provider credentials without impacting applications grant scoped access to models through workloads and keys this separation makes it easy to switch providers, compare models, and manage access as your usage evolves viewing existing providers to view and manage provider credentials go to model management in the left hand navigation by default, you will land on the providers tab this page shows all previously created providers, including display name provider creation date creating a new provider to create a new provider click add provider in the upper right enter a display name the name is purely for your reference use a naming convention that makes it easy to identify what the provider credential is used for, such as the provider, environment, or purpose select an ai provider from the list of supported providers once a provider is selected, the required authentication fields for that provider will appear provider specific configuration each provider requires different information for example openai typically requires an api key and optionally a custom base url aws bedrock requires multiple fields, such as aws access key id, aws secret access key, region, and other provider specific settings fill in all required fields for the selected provider click save provider the credential will now appear in your credentials list and can be used when configuring models deleting a provider you can delete a provider at any time however, it is important to understand the impact deleting a provider will remove the provider credential from amberflo automatically remove all models that depend on that provider prevent any workloads or access keys from using those models if a provider is in active use, deleting it will immediately break access to the associated models before deleting a credential, make sure it is no longer referenced by models you still need what’s next after creating a provider, the next step is to define one or more models that use that provider credential models determine which specific llms are available through the gateway and are what workloads ultimately gain access to continue to the models section to configure models for your provider