Skip to main content

Prebuilt crypto datasets for analytics: what Goldsky offers

Every prebuilt Goldsky dataset for EVM chains, Solana, Hyperliquid, and Polymarket, with what each holds and how to access it: Turbo pipelines, the Feeds API, or free SQL on CryptoHouse.

Prebuilt crypto datasets for analytics: what Goldsky offers cover image
Written by

Goldsky publishes prebuilt, decoded onchain datasets for EVM chains, Solana, Hyperliquid HyperCore, and Polymarket. A data team can read them in three ways. It can stream a Turbo pipeline into its own database or warehouse. It can call the Feeds REST API for wallet and chain data with no pipeline to run. It can also query Solana, Ethereum, and Base for free at CryptoHouse. Each dataset keeps the same schema on every chain, so one query works on several chains. The sections below list what each dataset holds, which chains it covers, and how to access it.

EVM datasets: blocks, transactions, logs, traces, and token transfers

The Turbo EVM source (opens in a new tab) offers 8 dataset types for most EVM chains. They are raw_blocks, raw_transactions, raw_logs, and raw_traces, plus the curated erc20_transfers, erc721_transfers, erc1155_transfers, and native_transfers. Every chain uses the same field names. A pipeline written for Base's erc20_transfers runs on Polygon or Arbitrum with only the dataset name changed.

Turbo Pipelines cover 114 of the 150 chains on goldsky.com/chains on November 16, 2026. That is 99 EVM chains and 15 other chains, including Solana, Bitcoin, Sui, Starknet, Stellar, NEAR, and Hyperliquid.

A team that needs decoded transfer history on several EVM chains does not write one indexer per chain. It points Turbo at the same dataset name on each chain and unions the results. A team that still runs Mirror pipelines can read the Turbo vs. Mirror comparison before it moves.

Solana datasets: transfers, balances, stablecoins, RWAs, and DEX swaps

Turbo's Solana source starts from genesis, according to the Turbo sources overview (opens in a new tab). The Solana source (opens in a new tab) has solana.blocks, solana.transactions, solana.instructions, solana.token_transfers, solana.native_balances, solana.token_balances, and solana.rewards. Three curated datasets sit on top of the transfer stream.

A stablecoin or RWA query on Solana does not start with a mint-address registry. The dataset already groups mints by currency or category, so a team filters on currency = 'USD' or category = 'treasury'. Index Solana data into Postgres shows a full Solana pipeline. USDC transfer tracker and Stablecoin issuance data use the stablecoin data.

Hyperliquid HyperCore and Polymarket

Goldsky curates 5 Hyperliquid HyperCore datasets (opens in a new tab): hypercore.blocks, hypercore.fills (with liquidation detail), hypercore.trades, hypercore.transactions, and hypercore.misc_events. Goldsky support gives access to them on request.

Polymarket (opens in a new tab) has 4 curated Turbo datasets: order_filled, orders_matched, user_balances, and user_positions. It also has 3 Feeds endpoints, Activity, Positions, and Balances, that return the same data over REST. Polymarket started markets on its Protocol V2 contracts on November 9, 2026. The current Turbo versions of the 4 datasets include V2 data.

A team that needs positions for a few wallets can call the Feeds Positions endpoint and skip the backfill. A team that needs every fill or every position change streams the Turbo dataset. Two worked examples use these datasets: Real-time prediction market odds data and How to index prediction market outcomes.

Feeds: wallet, token, and block data with no pipeline

Feeds is a REST API. The Feeds overview (opens in a new tab) lists Balances and Transfers on Solana and 9 EVM chains. The EVM chains are Ethereum, Base, Arbitrum One, Optimism, Polygon, BNB Smart Chain, Robinhood Chain, Arc, and Elysium. Tokens covers the same 9 EVM chains. Blocks and Chain head cover 21 chains. Every Goldsky plan, Starter included, gets a Feeds key. A caller with no account can pay per request in USDC over x402.

bash
curl "https://edge.goldsky.com/data/feeds/wallets/transfers?address=0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045&chains=base&direction=in&page_size=1&key=$GOLDSKY_FEEDS_API_KEY"

The call above returns the wallet's most recent inbound transfer on Base, with the USD value attached. A notification or reconciliation feature that needs one wallet's recent activity does not need its own database.

CryptoHouse: free SQL over Solana, Ethereum, and Base

Goldsky and ClickHouse run CryptoHouse (opens in a new tab), a free SQL service at crypto.clickhouse.com (opens in a new tab). It covers Solana blocks, transactions, token transfers, block rewards, accounts, and tokens, with similar datasets for Ethereum and Base.

A team can test a query in the browser before it deploys a Turbo pipeline. Historical blockchain dataset download covers the full block history behind these chains, and how Goldsky checked it against Solana's public RPC.

How to stream a dataset to your own database

This is the stablecoin transfer pipeline from the Solana Stablecoin Transfers docs (opens in a new tab), as published. It streams Solana stablecoin transfers to Postgres, with amounts scaled to token units.

yaml
name: solana-stablecoin-transfers
resource_size: m

sources:
  transfers:
    type: dataset
    dataset_name: solana.stablecoin_transfers
    version: "1.0.0"

transforms:
  scaled_transfers:
    type: sql
    primary_key: id
    sql: |
      SELECT
        id,
        signature,
        block_slot,
        block_timestamp,
        from_owner,
        to_owner,
        token_mint_address,
        symbol,
        currency,
        amount,
        decimals,
        CAST(amount AS DOUBLE) / POWER(10, decimals) as amount_scaled,
        transfer_type,
        _gs_op
      FROM transfers
      WHERE status = 1

sinks:
  postgres_transfers:
    type: postgres
    from: scaled_transfers
    schema: public
    table: solana_stablecoin_transfers
    secret_name: MY_POSTGRES
    primary_key: id

Every dataset in the catalog below uses the same source and transform structure. Only dataset_name and the transform fields change. Postgres is one sink. Turbo also writes to ClickHouse, Kafka, S3, webhooks, and more, listed in Turbo supported sinks (opens in a new tab).

Tokenized treasury onchain data applies this pattern to tokenized treasury funds. A team that needs to write data back onchain uses a different product, described in Custom oracle infrastructure.

Dataset catalog

DatasetWhat it holdsChainsAccess
EVM raw datasetsBlocks, transactions, logs, traces99 EVM chains with TurboTurbo pipeline
EVM curated transfersERC-20, ERC-721, ERC-1155, and native transfersMost EVM chains with TurboTurbo pipeline
Solana core datasetsBlocks, transactions, instructions, token transfers, balances, rewardsSolana, from genesisTurbo pipeline
Solana stablecoin transfersTransfers of 45 stablecoin mints in 14 currenciesSolanaTurbo pipeline
Solana RWA transfersTransfers of tokenized stocks, ETFs, treasuries, and metalsSolanaTurbo pipeline
Solana DEX swapsNormalized swaps across major AMM and bonding-curve programsSolanaTurbo pipeline
Hyperliquid HyperCoreBlocks, fills, trades, transactions, misc eventsHyperCore, from January 1, 2026Turbo pipeline, access on request
PolymarketOrder fills, matched orders, user balances, user positionsPolymarket (Polygon)Turbo pipeline or Feeds
Feeds wallet dataToken balances and transfer history, with USD values9 EVM chains and SolanaREST API, key or x402
Feeds chain dataBlock headers and chain head21 chainsREST API
CryptoHouseFree SQL over blocks, transactions, token transfers, accounts, tokensSolana, Ethereum, BaseBrowser, free

FAQ

Does a team need to run a pipeline to use these datasets?
No. Feeds answers wallet, token, and block questions over REST. CryptoHouse answers ad hoc SQL questions on Solana, Ethereum, and Base for free. A pipeline is for teams that want the rows in their own database.

Which dataset has ERC-20 token transfers?
<chain>.erc20_transfers on the Turbo EVM source. Each chain has one, with the same field names.

How far back does Solana data go on Turbo?
To genesis. A Solana source starts at the slot set in start_block, so a backfill can start at slot 0.

Can a team try a dataset before it pays for a pipeline?
Yes. CryptoHouse is free for Solana, Ethereum, and Base. Every plan, Starter included, gets a Feeds key.

How does a team filter the DEX swap dataset to one program?
Set filter.program_ids to the program address and options.dex to its label. The Solana DEX Trades docs page lists both.

© Endless Sky Inc. All rights reserved.

PrivacyTermsSecurity

System status