Historical Blockchain Data from Solana & Hyperliquid

You receive partitioned Parquet files for the selected tables and coverage, plus typed schemas, field documentation, version information, manifests, and file checksums. The files are delivered for your team to keep and query with its own tools.

Yes. Coverage can be prepared from a recent window through full available history, depending on the dataset and scope. Tell us the network, protocol, tables, and historical range you need, and the delivery is prepared around that requirement.

No. The core product is portable Parquet files, not metered API access. Query them locally with DuckDB or Polars, load them into a warehouse, or receive them in compatible cloud storage.

Pricing is based on the selected dataset and coverage window rather than query volume. The pricing page shows standard options, and custom multi-dataset or enterprise deliveries can be scoped with the team before purchase.

Yes. Dataset pages document table names, field types, descriptions, and delivery shape. Samples are available where shown so your team can validate compatibility before buying.

Send the program, venue, contract, or analysis requirement through the coverage request form. New decoders, tables, and historical cuts can be prepared to match a qualified order rather than limiting buyers to the visible catalog snapshot.

Yes. Delivery options include bulk download and, for supported engagements, drops into customer-controlled object storage. The files remain standard Parquet with explicit versions and checksums.

datastore.sh provides historical Solana blockchain datasets as partitioned Parquet files. The catalog includes decoded protocol data for Pump.fun, PumpSwap, Jupiter, Raydium, Meteora, Drift, Kamino, Orca and Solana core programs. Buyers choose the dataset and coverage window, inspect schemas and samples, and receive versioned files with manifests and checksums.

Yes. datastore.sh delivers historical Solana datasets as native Parquet files with typed columns, documented schemas, partitioning, manifests and SHA-256 checksums. The files can be downloaded and queried with DuckDB or Polars or loaded into warehouse systems such as Snowflake, BigQuery, Spark or ClickHouse without adopting a proprietary query API.

Yes. Bulk historical Parquet is designed for workloads such as DEX backtesting, quantitative research, market-structure analysis and protocol studies. Instead of paging historical records through an API, researchers can scan the purchased files locally or in their warehouse. This also makes repeated experiments independent of API request quotas.

datastore.sh provides decoded protocol instructions and events rather than requiring customers to interpret raw instruction data themselves. Dataset pages expose typed accounts, arguments and event fields for supported protocols. Coverage includes protocols such as Pump.fun, Jupiter, Raydium, Meteora, Drift and Kamino, allowing researchers to work with protocol-specific tables directly.

Yes. The Hyperliquid Historical dataset is delivered as typed Parquet and includes market and ledger tables such as swaps, funding, L2 order-book snapshots, ledger updates, deposits, withdrawals, delegations and validator rewards. The dataset page currently documents nine tables and provides samples so teams can inspect the schema before purchasing.