0xArchive comparison · Nodes and indexers
A node provider sells you access to the chain: RPC calls, current state, and streams of what happens from now on. An onchain indexer turns blocks into queryable tables of transfers, contract events and balances. A market data API such as 0xArchive sells the exchange's past: order books, trades, funding, open interest and liquidations for Hyperliquid and Lighter, recorded and served by time range.
All three touch the same venues. They answer different questions, and the common setup pairs two of them.
Which one answers your question
| Your question | Use | Why |
|---|---|---|
| Send a transaction, read current state | Node or RPC provider | It talks to the chain directly |
| Wallet transfers, token balances, contract events | Onchain indexer | It turns blocks into tables |
| What the book looked like at 14:32 last March | Market data API | It recorded the book then |
| Funding, open interest and liquidations over time | Market data API | Time series, already built |
| Fastest live view of your own orders | Your own node | Nothing sits in between |
A node records the market only from the day you start it
Hyperliquid publishes a non-validating node anyone can run. With its write flags on, it saves trades, fills, every order status and every raw book change to disk as they happen. That is the most direct live feed there is.
It is also a machine to keep alive. Hyperliquid's spec for it is a 16-core server with 128 GB of RAM. At default settings its logs grow by about 100 GB a day. Its local info server answers no historical time-series queries, so last month's book is whatever your own disk kept, if the node was running then.
Hosted RPC and node providers take the machine off your hands. What you get is still the chain's view: current state and streams going forward, not a dated archive of the order book.
An indexer reads the chain, not the order book
Indexers are built for what lands on the chain as state: transfers, deposits, contract calls and balances, usually queried with SQL or GraphQL. On Hyperliquid that fits HyperEVM contracts and account flows well.
A resting order that is placed and cancelled two seconds later changes the book without ever becoming a balance or a transfer. Rebuilding a book from indexed tables means replaying every order event in sequence, which is the job a market data archive has already done.
A market data API sells the exchange's past
0xArchive has recorded Hyperliquid's order book since 15 April 2023 and serves it with trades, funding, open interest and liquidations by venue, market and UTC window. From 10 March 2026 it also keeps the order-level book (L4), with the order ID and wallet behind every resting order. On Lighter it keeps trades from January 2025 and the order-level book (L3) from March 2026.
It is not a node. It does not send transactions, and its live streams reach you after a node of your own would see the same change.
Run a node for live trading and an archive for research
The node gives the lowest-latency view of the market now. The archive holds the years before the node existed, plus the replays a backtest needs.
If you run your own Hyperliquid node, 0xArchive also sells node peering: an allowlisted gossip peer in Tokyo for $200 per node address a month, flat. It is separate from the data plans and needs no 0xArchive account.
Route references
Use the REST reference for order books, trades, funding and liquidations, the Hyperliquid L4 reference for order-level books with wallets, the Lighter L3 reference for Lighter's order-level book, and venue coverage for each market's start dates.
For current API availability and plan details, use the Hyperliquid data API, data catalog, pricing, and status.