Create LLM credential

Create a new LLM credential.

Body Params

Schema for creating an LLM credentials.

Required fields vary by creds_type:

  • API_KEY: Requires api_key. Optionally base_url for OpenAI-compatible providers.
  • AWS: Requires aws_access_key_id, aws_secret_access_key, and aws_region.
  • AZURE: Requires azure_endpoint, api_version, and (api_key OR service principal creds).
  • GCP: Requires api_key OR (gcp_project, gcp_location, gcp_service_account).

string
required
length ≥ 1

Display name for this LLM credentials.

string
enum
required

LLM credentials type. Determines which other fields are required.

Allowed:
LLMProvider | null

The LLM provider these credentials authenticate against.

Optional. Required for API_KEY creds_type to enable creation-time validation.
For AWS/AZURE/GCP, provider is inferred from creds_type if not specified.

API key for LLM providers. Required for API_KEY auth.

Custom base URL for OpenAI-compatible providers (e.g., DeepSeek, Together, Groq).

AWS access key ID. Required for AWS auth.

AWS secret access key. Required for AWS auth.

AWS session token (optional, for temporary credentials).

AWS region for Bedrock (e.g., 'us-east-1'). Required for AWS auth.

Azure OpenAI endpoint URL (e.g., 'https://myresource.openai.azure.com/').

Azure API version (e.g., '2024-10-21'). Required for AZURE auth.

Azure AD client ID for service principal auth.

Azure AD client secret for service principal auth.

Azure AD token URL (contains tenant ID).

Google Cloud project ID. Required for GCP service account auth.

Google Cloud location (e.g., 'us-central1'). Required for GCP service account auth.

object | null

GCP service account JSON credentials. Required for GCP service account auth.

Responses

Language
Credentials
URL
LoadingLoading…
Response
Choose an example:
application/json