Overview

Streams gives you real-time and historical blockchain data — but just like blockchains, the size and shape of that data can vary. This guide will help you estimate potential costs and optimize your setup.

What Impacts Cost?

You’re billed based on data volume delivered, measured in GB (after gzip compression).

Key Factors

• Dataset Type — For example: block, transactions, block_with_receipts,  or logs. Each dataset type differs in size and complexity.
• Blockchain — Some blockchains generate more data per block. For example, Solana produces significantly more data than lighter chains like Polygon or Gnosis.
• Block Range — Older blocks often contain less data compared to recent ones.
• Block Activity — High-traffic blocks (e.g., during mints or MEV events) cost more to stream than quiet blocks.
• Filters — Apply filters to reduce the data volume before billing — you only pay for the data you keep.
• Compression — Streams are automatically compressed using gzip, which typically reduces payload size by 60–85% or more.

Learn More about Compression

Check out our Reduce Streams Costs with Data Compression to learn how to optimize your Streams setup.

Simple Way to Estimate

1. Choose a representative time window

Pick a recent hour or day that reflects your expected usage.

2. Set up a test stream

Use the Streams UI with your desired dataset, filters, and range.

3. Check the compressed output size

Look for the gzip-compressed size in the test output — that’s what you'll pay for.

4. Estimate your monthly cost

Example:

20 MB/hour × 24 hours/day × 30 days = ~14.4 GB/month

Cost-Saving Tips

Filter your data

Only pay for what you keep — even a basic filter can drastically cut volume.

Enable gzip compression

Compression is automatic and can reduce size by 60–85% or more.

Test historical vs. recent data

Querying older blocks may be cheaper due to smaller data size.

Use null to your advantage

If your filter returns null, you're not billed for that block.

We ❤️ Feedback!

If you have any feedback or questions about this documentation, let us know. We'd love to hear from you!