OLAPDBGet new results
Profiles

Every database, strengths and weaknesses

What each system is, what it does well, where it struggles, and the workloads it suits. Benchmark figures refer to the ClickBench runs on the benchmarks page.

ClickHouse

Apache 2.0

Open-source columnar database · self-hosted or ClickHouse Cloud

Best overall for real-time analytics

  • +Lowest ClickBench hot-run score in both groups: 1.03 on c6a.4xlarge, 1.06 in the managed group.
  • +Smallest on-disk footprint of the self-hosted engines: 9.4 GB for 100M rows, under half of DuckDB or StarRocks.
  • +Sustains 2.4–3× StarRocks' concurrent throughput on the same bare-metal hosts.
  • +Built for continuous inserts: Kafka and S3 table engines, async inserts and materialized views that update on write.
  • +Open source under Apache 2.0, with the same engine behind its managed cloud, so there is no lock-in.
  • −Joins between very large tables need more care than in a warehouse built around them.
  • −Updates and deletes are possible but are not its strength; it favours append-mostly data.
  • −Schema design (sort keys, partitions) matters a lot and takes some learning.
Read the profile →

DuckDB

MIT

Embedded, in-process analytical database

Best single-user, in-process engine

  • +Nothing to run: an embedded library with no server to operate.
  • +Best cold-run score of the self-hosted group (1.31) and the fastest bulk load (126 s).
  • +Reads Parquet, CSV and JSON in place, locally or on object storage.
  • +Excellent SQL ergonomics for analysts.
  • −One process, one writer: not built to serve many users or applications at once.
  • −About 2× ClickHouse's hot-run score on identical hardware.
  • −Took 20.5 GB on disk for the dataset ClickHouse stores in 9.4 GB.
  • −No continuous ingestion; data arrives in batches.
Read the profile →

BigQuery

Proprietary

Serverless data warehouse on Google Cloud

Easiest to start, hardest to predict the bill

  • +No infrastructure to size or tune.
  • +Most compact storage reported in the managed group (8.8 GB).
  • +Deep integration with the rest of Google Cloud.
  • −Roughly 370 ms even for a full COUNT(*), which ClickHouse answers in about 2 ms.
  • −On-demand pricing per byte scanned makes the cost of a busy dashboard hard to forecast.
  • −Slowest bulk load in the managed group except Redshift (777 s).
  • −Google Cloud only.
Read the profile →

Snowflake

Proprietary

Cloud data warehouse

Strong warehouse, slow for interactive analytics

  • +Mature governance, security and data-sharing features.
  • +Separation of storage and compute with simple warehouse sizing.
  • +Fast bulk load in ClickBench (67 s).
  • −Worst hot-run score of the managed group: 5.66.
  • −Bigger warehouses barely help: Medium, Large and X-Large all take about 60 s for the full query set.
  • −Credits are billed per second with a 60-second minimum each time a warehouse resumes.
  • −Closed source and cloud-only.
Read the profile →

Amazon Redshift

Proprietary

Data warehouse on AWS

Solid in AWS, slow to load

  • +Second-best hot score of the managed group (2.77).
  • +Tight integration with S3, Glue and the rest of AWS.
  • +Serverless option removes cluster sizing.
  • −1,889 s to load 100M rows: over 50× ClickHouse Cloud.
  • −30.3 GB on disk, 3× ClickHouse Cloud for the same data.
  • −Published ClickBench results are from 2022; newer runs have not been submitted.
  • −AWS only.
Read the profile →

Databricks SQL

Proprietary service on open formats (Delta Lake, Spark)

Lakehouse SQL warehouse

Best if your data already lives in the lakehouse

  • +One platform for ETL, ML and SQL over open table formats.
  • +Fast bulk load in ClickBench (53 s on Medium).
  • +Compact storage (10.2 GB).
  • −Hot-run score of 5.16, about 5× ClickHouse Cloud.
  • −Fixed per-query overhead: around 270 ms for a full COUNT(*).
  • −DBU pricing makes it the most expensive per query in our estimate.
Read the profile →

StarRocks

Apache 2.0

Open-source MPP analytical database

Capable open-source alternative, behind on speed and storage

  • +Open source under Apache 2.0.
  • +Cost-based optimizer that handles complex joins well.
  • +Stream Load and Routine Load for real-time ingestion from Kafka.
  • +MySQL-compatible protocol works with existing tools.
  • −Hot-run score of 2.9 on c6a.4xlarge, almost 3× ClickHouse.
  • −Slowest bulk load of the self-hosted group (489 s).
  • −A third to two-fifths of ClickHouse’s concurrent queries per second on the same bare-metal hosts.
  • −More components to operate (frontend and backend nodes).
Read the profile →