I need a warehouse that can query JSON and parquet data

If you need a warehouse that can query JSON and Parquet directly, you have several good options depending on whether you want SQL analytics, cloud-managed, or open-source.

Best options

1. Snowflake

• Can query JSON via VARIANT
• Can query Parquet in external/internal stages
• Strong for analytics, scalability, and low ops
• Good if you want a managed warehouse

2. Google BigQuery

• Excellent support for JSON
• Can query Parquet in external tables and load jobs
• Serverless, fast to start with
• Great if you’re already in GCP or want minimal infrastructure

3. Databricks SQL / Delta Lake

• Very good for Parquet
• Can handle JSON as well
• Best if you want a lakehouse approach and more data engineering flexibility
• Strong with large-scale semi-structured data

4. Amazon Redshift

• Supports JSON and Parquet
• Can query data in S3 via Redshift Spectrum
• Good if you’re in AWS and want a warehouse with external data access

5. DuckDB

• Can query JSON and Parquet directly
• Lightweight, local, and extremely convenient
• Not a traditional warehouse, but great for ad hoc analytics and prototyping

Quick recommendation

• Best managed warehouse: Snowflake
• Best serverless: BigQuery
• Best open-source / local analytics: DuckDB
• Best lakehouse: Databricks

Example of what “querying JSON and Parquet” usually means

• JSON:
  • nested fields
  • semi-structured event data
• Parquet:
  • columnar analytics files
  • common in data lakes

If you want one that can query files without loading them first

Look at:

• Snowflake external tables / stages
• BigQuery external tables
• Redshift Spectrum
• DuckDB directly on files

If you tell me:

1. cloud provider you use,
2. data size, and
3. whether you want managed vs open-source,

I can recommend the best fit more precisely.