Hybrid tables leverage a row store as the primary data store to provide excellent operational query performance. When you write to a hybrid table, the data is written directly into the rowstore. Data is asynchronously copied into object storage in order to provide better performance and workload isolation for large scans without impacting your ongoing operational workloads.

Some data may also be cached in columnar format on your warehouse in order to provide better performance on analytical queries. You simply execute SQL statements against the logical hybrid table and Snowflake's query optimizer decides where to read data from in order to provide the best performance.

Reference - https://quickstarts.snowflake.com/guide/getting_started_with_hybrid_tables/index.html?index=..%2F..index#0