apache/iceberg-python

Health v2 · maintenance only

Adoption not part of the score

1124 stars · 572 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

PyIceberg is a Python implementation of the Apache Iceberg table format specification, providing programmatic access to Iceberg table metadata and data without requiring a JVM. It supports multiple catalogs (Hive, AWS Glue, REST, SQL) and file IO backends (S3, GCS, ADLS) with integrations for PyArrow, Pandas, DuckDB, Ray, Polars, and other engines.

Use cases

• read and write Apache Iceberg tables from Python without a JVM
• query Iceberg tables with pandas, DuckDB, or Polars
• manage Iceberg table metadata and catalogs programmatically
• connect to AWS Glue, Hive, or REST catalogs for lakehouse tables
• load Iceberg table data from S3, GCS, or Azure object stores
• append or overwrite data files in an Iceberg table

When to choose

• you work with Iceberg lakehouse tables in a Python data stack
• you want JVM-free access to Iceberg tables
• you need to integrate Iceberg with PyArrow, DuckDB, Ray, or Polars

When to avoid

• you need a full distributed query engine rather than table access
• your tables use a format other than Iceberg (e.g. Delta Lake or Hudi)
• you require features of the Java Iceberg runtime not yet implemented in Python

Facets

library · maturity active

database data-science etl databases big-data analytics python cross-platform apache-iceberg table-format lakehouse pyarrow object-storage catalog data-engineering

Member repositories

For agents

markdown · JSON · MCP: product_card(name="apache/iceberg-python")