What Hyperliquid is

Hyperliquid is a layer 1 blockchain written from first principles, in its own documentation’s words ^[4]. It has two execution components. HyperCore, the original trading engine. HyperEVM, the EVM-compatible layer. The docs are explicit about the relationship. The HyperEVM is not a separate chain ^[4]. It is secured by the same HyperBFT consensus as HyperCore ^[4]. Both share one unified state ^[4].

One caveat the issuer carries itself. The HyperEVM is in the alpha stage ^[4]. That label is the issuer’s, kept visible here.

A launch date this page does not state

Many sources give a Hyperliquid launch year. None of the banked primary ones do. The official RPC returns no block-zero object. A second provider serves the genesis hash with a zero timestamp, a null header. The captured documentation states no launch date. This page will not print a date the primary record does not carry. The read date for that negative is 11 September 2026.

How consensus works

Hyperliquid uses HyperBFT, a custom consensus inspired by HotStuff and its successors ^[4]. Like most proof-of-stake chains, blocks are produced by validators in proportion to the native token staked ^[4].

The active set is transparent. It is determined by the top twenty-seven validators by stake ^[4].

Latency figures need careful labels. The docs state a 0.2 second median end-to-end latency for the order path ^[4]. That is order round-trip latency, issuer-stated. It is not a block time. HyperCore carries one-block finality inherited from HyperBFT ^[4].

A block cadence exists as observation only. Ten consecutive blocks spanned 16 seconds in one window, about 1.6 seconds per block ^[5]. Single observation, small sample, labeled as such. The documentation states no block time.

The native token

HYPE is the native gas token on the HyperEVM ^[4]. It carries 18 decimals on both mainnet and testnet ^[4]. The chain ID for the EVM side is 999, matching the documentation value ^[5].

How fees work, with a difference

HyperEVM runs the Cancun hardfork without blobs, with EIP-1559 enabled ^[4]. Base fees are burned, as usual. Then the difference. Unlike most EVM chains, priority fees are also burned, because the chain uses HyperBFT consensus ^[4]. Burned priority fees go to the zero address ^[4].

No validator tip leg exists. Observed fee medians here carry no tip component by design. Comparing them against tip-inclusive medians from other chains without this note would mislead, so the note rides with the figures.

Fees, measured

The conditions metric, dated 11 September 2026. Median base fee of 0.1038 gwei times 67,757 gas, which is 0.0000070331 HYPE, about $0.00056 ^[5] ^[8]. The base fee sits at a 0.1 gwei floor on both providers sampled ^[6].

One method deviation, documented rather than hidden. This library’s set-wide rule measures simple-transfer gas as the median of transfers under 60,000 gas. On HyperEVM that subset is empty. The lowest observed USDC transfer used 62,159 gas. The simple-transfer plateau is 67,757, an exact value repeated across five or more transfers on both providers. The plateau stands in, and the deviation is recorded here ^[5].

The observed metric. Fifty sampled USDC transfers. The median paid was 0.0000735041 HYPE, about $0.0058 ^[5] ^[8]. Batch-inclusive, like every observed metric in this library. The second provider’s independent window returned about $0.0052, with an identical simple-gas plateau ^[6].

The checking trail, stated. The official endpoint served the primary window ^[5]. An independent node served the full re-pull, 26 transfers, with a floor-exact 0.1 gwei median base against the official 0.1038 ^[6]. That is a 3.8 percent gap between providers, recorded rather than averaged. The two nodes agree exactly on the 67,757 plateau and on the observed minimum base ^[6]. No explorer gas oracle exists for this chain. The USD conversion uses a same-session provider price, a conversion aid, not a chain fact ^[8].

One route note on the docs. The Hyperliquid documentation host serves a script shell to plain fetches. The markdown source behind each page is exposed by the host’s own convention, and that markdown is what the captures hold ^[4].

The stablecoin economy

One provider’s chains table lists a Hyperliquid L1 row at $6,939,629,935.94 on 11 September 2026 ^[7]. That provider treats the core and the EVM as one chain. That matches the unified-state architecture the docs describe ^[4].

The roster

One issuer-listed dollar stablecoin runs natively on the HyperEVM. The address is issuer-doc verbatim, from Circle’s current table, captured fresh on 11 September 2026 ^[1]. Two independent nodes decoded the contract identically, name and decimals ^[5] ^[6].

Token	Address, as the issuer publishes it
USDC, Circle ^[1]	`0xb88339CB7199b77E23DB6E890353E22632Ba630f`

The absences carry equal weight. Tether’s protocols page contains zero mentions of Hyperliquid ^[3]. Circle’s EURC table carries no Hyperliquid row ^[2]. No other banked issuer rows exist for this chain in the research set. Wrapped HYPE assets are wrapping vehicles, not stablecoin roster entries.

Where this page sits

The chains hub holds the other networks. The USDC page holds the token this chain’s roster carries. The Arbitrum page profiles a chain with the opposite fee treatment, where tips exist. The layer 1 versus layer 2 page uses this chain as its layer 1 baseline. Every figure above carries its date.

Fact	Value	As of	Source
Architecture	Layer 1, HyperCore plus HyperEVM, one unified state	2026-09-11	C-42
Consensus	HyperBFT, active validators are the top 27 by stake	2026-09-11	C-42
Order latency (issuer-stated)	0.2 second median end to end	2026-09-11	C-42
Native token	HYPE, 18 decimals, gas on HyperEVM	2026-09-11	C-42
Fee model	EIP-1559, base fee and priority fees both burned	2026-09-11	C-42
Conditions USDC transfer fee	$0.00056, at the 0.1 gwei base floor	2026-09-11	C-43
Observed median USDC transfer fee	$0.0058, batch-inclusive	2026-09-11	C-43
Stablecoins on chain (provider)	$6,939,629,935.94	2026-09-11	C-10