GoKawiilPolars 2.0 Release Candidate Makes Streaming Engine the Default
Polars has shipped the first release candidate for version 2.0, with the final release expected in coming weeks. The major version bump primarily changes the default execution engine for LazyFrame queries from the in-memory engine to the streaming engine, alongside other default setting changes and removal of legacy design constraints.
GoKawiil's interpretation of the reporting above, not reported fact.
Switching the default engine means most users will see substantial memory and speed gains—Polars estimates roughly 5x faster performance—without changing their code. However, the streaming engine doesn't guarantee row order for operations like joins and group_by, so users relying on deterministic ordering must explicitly opt in with maintain_order=True or revert to the old engine, making this a breaking change disguised as a routine upgrade.
- Streaming engine becomes the default for LazyFrame.collect() in Polars 2.0
- Row order is no longer guaranteed by default for joins, group_by, and unpivot operations
- Users can opt into observable order or switch back to the in-memory engine via configuration
Python Polars: The Definitive Guide (O'Reilly book) — With Polars 2.0 making the streaming engine the default for LazyFrame queries, it's a great moment to get properly fluent in lazy evaluation and expressions. This O'Reilly guide by Jeroen Janssens and Thijs Nieuwdorp walks through the Polars API from the ground up, so migration notes actually make sense instead of feeling like guesswork.
See Python Polars: The Definitive Guide (O'Reilly book) on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.Source: pola.rs, 2026-09-03
Published there as: “Pre-Release of Polars 2.0”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.