BitqueryBitqueryData Store
Self-serve. No sales call, no API key, no quota.

Historical blockchain data, delivered as files.

Buy bulk on-chain datasets from Bitquery as partitioned Parquet. See the full schema, download a free sample, pay by card or bank transfer, and receive signed download links by email.

datasets
17
on 1 network
documented columns
2,345
across 34 tables, every one typed
Featured datasets

Start with these

View all

Coverage across

Need a chain that isn’t listed? Ask for a custom scope

How it works

From landing page to downloadable files in four steps

  1. 01

    Pick a dataset

    Every dataset page shows the exact tables, columns and types you get. No sales call to find out what is inside.

  2. 02

    Download the free sample

    Real records for every table, straight from our public sample bucket. Load them into DuckDB or pandas and check the fit before you pay anything.

  3. 03

    Pay the fixed price

    Full history, or a shorter window where a dataset offers one. Prices are shown up front. Pay by card or bank transfer on the invoice page.

  4. 04

    Receive download instructions by email

    After you pay, our team prepares your export and emails you signed download links.

The delivery contract

Files you can trust in a pipeline

  • Stable layout. chain / table / file per block range or day. The same layout as our public sample bucket, for every dataset.
  • Manifest and checksums. Every delivery ships a JSON manifest with a sha256 per file, coverage window and lineage.
  • Files are never overwritten. Schema changes never alter files already delivered. What you bought stays as it was.
  • Links re-issued on request. If a signed link expires before you download, write to sales@bitquery.io and we reissue access for the same order at no cost.
s3://bitquery-blockchain-dataset/
  solana/dex_trades/pumpfun/
    390740000_390740049.parquet
    390740050_390740099.parquet
    
  solana/pumpfun_creation_migrations/
    2026-07-01.parquet
  _manifests/
    pumpfun-swaps.v1.2.manifest.json

$ duckdb -c "SELECT count(*) FROM
    'solana/dex_trades/pumpfun/*.parquet'"

Frequently asked questions

What format is the data delivered in?

Apache Parquet with Snappy compression, flat records, one prefix per chain and table, files cut by block range or by day depending on the table. It loads directly into DuckDB, Polars, pandas, Spark, ClickHouse and Snowflake. CSV export is available on request for enterprise orders.

Is this the same as the Bitquery API?

No. The Data Store sells the files. There is no metering, no rate limit and no API key. If you need live or ad-hoc queries, the Bitquery GraphQL and streaming APIs are the right product, and we cross-link both ways.

Do I need an account to buy?

No. Checkout is guest checkout: an email for the receipt and a company name for the invoice. You receive the invoice and receipt by email, and we write to you when your files are ready.

How fresh is the data?

Published datasets are refreshed daily with T+1 latency. The coverage end date on each dataset page is the last date included in a purchase made today.

Can I get a refund?

Completed deliveries are not refundable. If we fail to deliver, or the delivery is the wrong listing, in a faulty format, or has proven gaps, we correct it within 7 business days of written notice. See section 9 of the terms.

What if I need a custom scope?

Custom windows, extra tables, CSV output or cloud-bucket delivery are handled through the enterprise route. Fill in the contact form and a data engineer will scope it.

Need live or ad-hoc queries instead?

The Data Store is for bulk historical files. For real-time streams, GraphQL queries and metered access, use the Bitquery API. Existing API customers get the same datasets here with no extra onboarding.

Explore the Bitquery API