How these docs work

One set of pages for every reader, from first-timers to protocol engineers. Each concept lives in exactly one place; everywhere else gives it a sentence and a link.

How the site is organised

• Top bar: five areas, by what you came to do: Learn, Trade, Build, Deploy, Reference.
• Tabs below it: the sections of your area. Open one to see every page in it.
• Back and Next, above every title and at the foot of every page, walk the reading order.

How each page reads

Every page opens by saying plainly what the thing is and what it does for you. A worked example with real numbers usually follows, then sections whose headings state their point, so you can scan the headings alone and get the gist.

• The line under the title is the whole page in one sentence.
• The numbers is a table of every parameter, formula and limit, in the same place on every page.
• “The official docs differ” boxes mark where measurement on chain contradicts the documentation.
• The last lines point you to where to go next, and the Next card takes you there.

Pages start simple and go deeper as you scroll. Stop when a page stops being useful: nothing further down is needed to understand what came before.

Levels

The pill under each title marks where the page’s weight sits, not who may read it.

Level	Assumes	What the page does
Foundation	Nothing	Says what a thing is and why it exists, with an analogy where one helps and a note on where the analogy breaks
Practitioner	You have placed a trade or read the foundation pages	Explains the mechanism with its real parameters, formulae and limits
Expert	You write code against the protocol	Gives wire formats, encodings and units, and the behaviour measured on chain

Where to start

• Trading: from a first order to margin, liquidation and professional practice.
• Building apps and front ends: the API, signing and orders, builder codes, then contracts on the HyperEVM.
• Deploying markets: HIP-3 perp dexes, HIP-4 outcome markets and HIP-1 tokens, and what each asks of you.
• Press and analysts: the structure, how to measure it and the names to get right.

No reason in mind yet? Start with Hyperliquid in five minutes.

Where the facts come from

Every number, field name, unit and limit comes from a written source. Sources rank:

1. Measurement on chain. Behaviour observed on mainnet or testnet, with transaction evidence and a sample size.
2. The official documentation. The authoritative wording of what the protocol is meant to do.
3. SDK code. Used for wire shapes and field names, never for economic parameters.

When they disagree, measurement wins, and the page says so in a caution with the sample size:

The official docs differ

The documentation says order actions and vault transfers sent from the HyperEVM through CoreWriter are delayed on chain for a few seconds. Measured on mainnet over 52 orders across 15 accounts and 14 markets, orders landed 0.35 to 0.51 s after their EVM block. Only vault transfers were delayed, by 4.05 to 4.10 s (6 observations).

Open questions are marked open. Measurements from the author’s own deployments give the method, not a company or system name.

Conventions

• Names are written exactly: HyperCore, never shortened; HyperEVM; HIP-3, not HIP3. The naming and style guide lists every one.
• Units are always stated: “1e8 per contract”, “tenths of a basis point”, “milliseconds”.
• Every link stays on this site and points at the page that owns the concept.
• Spelling is British.

For the whole picture on one page, start with Hyperliquid in five minutes. Every term used here is defined once in the glossary.