Polars
Amsterdam, Netherlands
What it is
Polars is an open-source DataFrame library for manipulating data with a typed, expressive API. Its Rust-based engine uses multiple threads and vectorized, columnar processing for parallel work on a single machine. A query optimizer selects execution plans, and the streaming API can process datasets larger than memory without retaining every result. Installation examples cover Python, Rust, and JavaScript. Polars supports CSV, JSON, Parquet, Delta Lake, Avro, Excel, Feather, Arrow IPC, databases, and cloud storage. It can read and write AWS S3, Azure Blob Storage, and Google Cloud Storage. Polars Cloud uses the same API for distributed queries, scaling from local execution to clusters on AWS; compute runs in the customer’s cloud environment. Polars On-Prem supports Kubernetes and bare-metal servers, including air-gapped Enterprise deployments. The open-source library is free under the MIT license. Polars Cloud offers a 30-day trial, and the listed paid starting price is $0.05.
For and against
For
- Free MIT-licensed open source
- Multithreaded Rust query engine
- Streaming for larger-than-memory datasets
- Broad file and cloud support
- Distributed cloud and on-prem options
Against
- On-Prem Self-Serve has one cluster
- On-Prem cluster limit is 1,024 cores
- On-Prem limit is 64 nodes
- Cloud trial starts after AWS connection
The verdict
Polars provides a code-first DataFrame workflow with local, streaming, cloud, and on-prem deployment paths. Its open-source library is free, while distributed services have stated limits and trial conditions.
Who it suits
Polars suits developers and data teams building code-based data transformation workflows across local, cloud, or on-premises environments. It supports users working in Python, Rust, or JavaScript.
Plans and prices
Polars (open source)
Open source data manipulation library
Read 7 Oct 2026 · pola.rs
On the record
- Cloud availability
- The Polars Cloud user guide says Polars Cloud is available on AWS.docs.cloud.pola.rs · read 7 Oct 2026
- Cloud service
- Polars Cloud runs distributed queries using the Polars API and can scale workloads from local execution to a cluster.cloud.pola.rs · read 7 Oct 2026
- Cloud storage
- The user guide says Polars can read and write data in AWS S3, Azure Blob Storage, and Google Cloud Storage.docs.pola.rs · read 7 Oct 2026
- Engine
- Its multithreaded query engine is written in Rust and uses vectorized, columnar processing for parallel execution.pola.rs · read 7 Oct 2026
- Integrations
- The homepage lists support for CSV, JSON, Parquet, Delta Lake, AVRO, Excel, Feather, Arrow, several databases, and cloud storage.pola.rs · read 7 Oct 2026
- Languages
- The site provides installation examples for Python, Rust, and JavaScript.pola.rs · read 7 Oct 2026
- Limits
- The On-Prem Self-Serve plan is limited to one concurrent cluster, 1,024 cores per cluster, and 64 nodes.docs.cloud.pola.rs · read 7 Oct 2026
- On-prem deployment
- Polars On-Prem supports Kubernetes distributions and bare-metal servers, including air-gapped Enterprise deployments.docs.cloud.pola.rs · read 7 Oct 2026
- Product
- Polars is an open-source DataFrame library for data manipulation, described as a fast single-machine data processing solution with a typed, expressive API.pola.rs · read 7 Oct 2026
- Query optimization
- Polars uses a query optimizer to determine an efficient execution plan for queries.pola.rs · read 7 Oct 2026
- Security
- Polars says Polars Cloud compute runs in the customer's cloud environment and that customer data does not leave that environment.docs.pola.rs · read 7 Oct 2026
- Streaming
- The streaming API can process datasets larger than memory without keeping all results in memory.pola.rs · read 7 Oct 2026
- Support
- Polars Cloud offers team support for technical questions, implementation guidance, and product feedback.cloud.pola.rs · read 7 Oct 2026
- Trial
- Polars Cloud offers a 30-day trial that begins when an organization connects its first workspace to AWS.docs.cloud.pola.rs · read 7 Oct 2026
