The following features have been added to Python UDFs during Preview:

The following features have been added to Python UDFs during Preview:

• Vectorized UDFs with Apache Arrow. You can now create vectorized Python UDFs using the Apache Arrow RecordBatch interface for improved performance. • Cloud Monitoring integration. Python UDFs now export metrics to Cloud Monitoring, including CPU utilization, memory utilization, and maximum concurrent requests per instance. • Container request concurrency. A new option, container_request_concurrency, is available for the CREATE FUNCTION statement. This option controls the maximum number of concurrent requests per Python UDF container instance. • New quotas. Python UDFs are now subject to new quotas on image storage bytes (10 GiB per project per region) and mutation rate (30 per minute per project per region). • Cost visibility. Python UDF costs can be seen in the external_service_costs column in the INFORMATION_SCHEMA.JOBS view and in the ExternalServiceCosts field in the Job API.