Use BigQuery as a Mage destination to load structured data into Google Cloud’s fully-managed, serverless data warehouse.

​
Configuration Parameters

When configuring BigQuery as a destination, provide the following parameters:

KeyDescriptionExample ValueRequired
path_to_credentials_json_fileAbsolute path to your Google Cloud service account credentials file (JSON). If Mage is running on a GCP instance, leave this as null to use the instance’s service account./path/to/credentials.json✅
project_idThe GCP project ID where your BigQuery dataset exists.example_project✅
datasetName of the BigQuery dataset where data will be written.example_dataset✅
location(Optional) The region of your BigQuery dataset. If not set, BigQuery may infer it automatically.us-west1❌
disable_update_column_typesIf false, Mage will alter existing BigQuery column types to match the schema when necessary. If true, existing types will not be modified.false (default)❌
use_batch_loadUse BigQuery load jobs instead of the streaming API. Recommended for better performance and cost efficiency.true (default)❌
credentials_infoAlternative to path_to_credentials_json_file. Directly embed your credentials JSON as a dictionary (useful for environment-based configs).See structure below❌

​
credentials_info Structure

If you’re passing credentials inline instead of a path, structure it like this:

auth_provider_x509_cert_url: str
auth_uri: str
client_email: str
client_id: str
client_x509_cert_url: str
private_key: str
private_key_id: str
project_id: str
token_uri: str
type: str

​
Optional Configuration

KeyDescriptionExample ValueDefault
skip_schema_creationIf true, Mage will not run a CREATE SCHEMA command. Useful if the schema already exists. See issue #3416.truefalse
lower_caseIf true, column names will be automatically converted to lowercase before writing to BigQuery.truetrue

​
Required Permissions

Your GCP service account or user must have the following permissions to use BigQuery as a destination:

bigquery.datasets.create
bigquery.jobs.create
bigquery.readsessions.create
bigquery.readsessions.getData

Additionally, the account must have the BigQuery Data Editor role on the target dataset:

  • Role: roles/bigquery.dataEditor