5 Best Hyperliquid RPC Providers of 2026

Hyperliquid runs two layers on one state: HyperCore, which holds the order books, and HyperEVM, which runs smart contracts. The official docs cap the public HyperEVM endpoint at 100 JSON-RPC requests per minute per IP, and REST calls share a weight budget of 1,200 per minute. A Hyperliquid RPC provider hosts nodes and APIs behind private keys, so a wallet, bot, or analytics job can read state and send transactions without those ceilings.

“Choosing a primary Hyperliquid RPC provider matters as much as choosing a backup, because any complex business needs fallback options. The five providers below can fill either role.”

— Kate Liftshits, CBDO of NOWPayments

Below is the ranked list of leading Hyperliquid node providers for 2026:

1. NOWNodes
2. QuickNode
3. Alchemy
4. Chainstack
5. dRPC

How We Ranked the Best Hyperliquid RPC Providers

Each provider was judged on its own pricing page, documentation, and Hyperliquid listing, all read on September 29, 2026. Where a number comes from vendor marketing, the text says so. Three criteria drive the ranking.

• Network Coverage and Node Types: which Hyperliquid layers and networks a vendor serves, plus shared, dedicated, and archive options.
• Performance and Reliability: written uptime terms, throughput limits, and how the vendor describes its infrastructure.
• Pricing and Support: free allowance, entry price, billing unit, and stated support response.

The table puts the five Hyperliquid RPC endpoint providers side by side.

Rates, supported chains, and limits change frequently. Verify current terms on each provider’s official site before committing.

NOWNodes

Best for: multi-chain teams that add Hyperliquid to an existing stack.

NOWNodes is one of the best RPC providers for teams that run many chains through a single account. It offers shared and dedicated nodes, including a hyperliquid rpc endpoint next to 120+ other networks. Its Hyperliquid page states 99.9% uptime and geo-balanced clusters with automatic failover. It suits teams that already route several chains through one vendor and want Hyperliquid on the same key.

Network Coverage and Node Types. The docs list the standard HyperEVM JSON-RPC methods plus 26 Hyperliquid info methods, including clearinghouseState, meta, and openOrders. Methods such as l2Book and userFills are absent, so order-book depth and fill history need another source. The product page names mainnet only and mentions no HyperCore streaming or gRPC for Hyperliquid. Dedicated nodes are sold next to the shared plans, one network per node.

Performance and Reliability. Paid plans carry no predefined RPS limits, so throughput follows cluster capacity. That describes a policy, not unlimited hardware. NOWNodes publishes no Hyperliquid-specific latency benchmark, which leaves p95 measurement to the buyer.

Pricing and Support. Start is free with 100,000 requests and one API key, but it excludes WebSocket. Pro adds WebSocket and 1 million requests, Pro Plus gives 10 million, Business 30 million, and Enterprise 100 million. Overage runs from €0.5 to €5 per 100,000 requests on Pro Plus and above. Every plan lists a 3-minute first response from support. The pages reviewed show no SOC or ISO reports, and the free allowance is the smallest of the five.

QuickNode

Best for: trading and analytics systems that need HyperCore data.

QuickNode reaches both HyperEVM and HyperCore through a single endpoint. Its documentation describes the broadest set of exchange-data interfaces on this list, and Enterprise customers can request SOC and ISO reports. It suits teams building order-book consumers and market-data pipelines.

Network Coverage and Node Types. HyperEVM splits into /evm, which keeps only recent blocks and prunes about every 12 hours, and /nanoreth, which adds archive, debug, trace, and WebSocket. HyperCore offers 71 of 73 info methods, six WebSocket streams, and 13 gRPC streams. HyperEVM and gRPC run on mainnet and testnet, while the exchange REST API is mainnet only.

Performance and Reliability. Rate limits grow with the tier, topping out at 500 RPS on Business. A contractual uptime SLA arrives only with Enterprise, together with SOC 1, SOC 2 Type 2, ISO 27001, and DORA reports on request. Sentry peering for lower latency is a paid add-on on every tier.

Pricing and Support. A one-month trial carries 10 million credits and 15 RPS, and the HyperCore endpoints stay closed until a paid plan. Build, the cheapest option that unlocks them, lists at $49 monthly or $34 on annual billing, with 80 million credits and a 24-hour support target. Credit weights per method make monthly bills hard to forecast. After the trial ends, there is no free tier to fall back on.

Alchemy

Best for: contract developers who want tooling beside RPC.

Alchemy offers HyperEVM RPC and WebSocket endpoints on mainnet and testnet inside a developer platform that spans 100+ chains. Its Hyperliquid page describes HyperEVM only. It suits teams that want webhooks and data APIs next to node access.

Network Coverage and Node Types. The chain page marks RPC, WebSockets, and the Debug API as available for Hyperliquid, while Trace API and gRPC carry a “request support” label. No HyperCore info or streaming endpoints appear. Single-tenant Dedicated Clusters exist as a product, but the Hyperliquid page does not mention them.

Performance and Reliability. Throughput is 25 RPS on the free plan and 300 RPS from Pay As You Go, with more capacity sold in $160 monthly blocks. Enterprise begins at 1,000 RPS and is the only tier with signed SLAs.

Pricing and Support. The free plan includes 30 million compute units a month, and usage beyond that costs $0.525 per million. Because every method has its own compute-unit weight, log and trace queries drain the quota faster than balance reads. Support targets are 48 hours on Free and 24 on Pay As You Go, and Premium packages respond in 12, 4, or 2 hours.

Chainstack

Best for: teams that want both layers and a route to dedicated nodes.

Chainstack hosts Hyperliquid for HyperEVM and HyperCore on mainnet and testnet, in full and archive modes. Its site states 99.99%+ uptime and lists SOC 2 Type II and ISO 27001. It suits teams that want a permanent free plan with room to grow.

Network Coverage and Node Types. Every Hyperliquid node exposes an /evm path and a /nanoreth path, and the latter includes protocol-level system transactions in block responses. Global nodes cover most needs, while a dedicated HyperEVM node starts from $803 a month on Chainstack’s page. Dedicated access goes through a request to support and takes one to two business days.

Performance and Reliability. Plan limits start at 25 RPS and reach 600 RPS on Business. Enterprise removes the cap and adds an uptime SLA, and Chainstack cites 99.9% for enterprise dedicated Hyperliquid nodes. Debug and trace start on Growth.

Pricing and Support. The Developer plan never expires and includes 3 million requests a month. Growth adds 20 million requests for $49, and dedicated nodes require Pro at $199. Paid plans get a 24-hour standard support response, and quicker tiers are add-ons priced at $1,000 or $10,000 a year. An archive call consumes two request units instead of one.

dRPC

Best for: teams that want an inexpensive backup endpoint.

dRPC routes traffic across several node operators and lists 130+ chains across 231 networks. Its Hyperliquid page covers HyperEVM on mainnet and testnet with HTTPS and WSS endpoints. It suits teams that keep a primary provider and need a cheap second path.

Network Coverage and Node Types. The listing marks archive access, and the Growth plan bundles a debug and trace API. HyperCore info and streaming do not appear. The free tier reaches every chain, though only through public nodes.

Performance and Reliability. The Growth plan lists a ceiling of 5,000 RPS, the highest published paid ceiling in this comparison, along with 99.99% uptime. Enterprise adds unlimited RPS and a service agreement. Since requests pass through a routing layer, response times depend on which operator answers, so a load test matters more than the spec sheet.

Pricing and Support. The free plan offers 210 million compute units a month, limited to 100 RPS. Growth bills $6 per million requests and prices archive and full-node calls alike. Enterprise contracts start at 300 million monthly requests. The pricing page promises priority support but names no response times.

How to Choose a Hyperliquid RPC Provider

Match the provider to what the application does every day, then read the fine print on limits. A plan that looks generous at idle can behave very differently during a volatile market.

• Pick the layer first. HyperEVM serves contracts and standard Ethereum tools, while HyperCore serves order books, positions, and fills.
• Check historical needs. Recent-block endpoints prune quickly, so old state and traces need an archive path.
• Compare limit types. RPS caps, credits, compute units, and request units all fail differently under a spike.
• Find the written uptime term. Most vendors here reserve signed SLAs for Enterprise contracts.
• Test support at your hour. Stated response targets range from 3 minutes to 24 hours.

On Hyperliquid, layer coverage matters more than headline RPS, since a larger plan cannot add an interface the vendor does not serve. Pilot two providers with production-representative traffic for a week before committing.

When a Hyperliquid RPC Provider May Not Be the Right Fit

Paid endpoints are not the answer to every Hyperliquid project. Four situations point elsewhere.

• Low traffic on a single IP. A script that stays under the public limit of 100 EVM requests per minute needs no paid plan.
• Latency measured in microseconds. Teams that want validator proximity or full control over node configuration run their own node or a dedicated sentry setup, since a shared endpoint adds network hops.
• Bulk historical analysis. Hyperliquid’s docs point large batch requests to its S3 bucket, and indexed data products suit aggregate queries better than raw RPC calls.
• Audit-driven procurement. If a contract demands SOC or ISO reports, that requirement alone removes several vendors from the list.

Hyperliquid RPC Provider Pricing Compared

Headline prices hide the unit each vendor bills in. The table prices a benchmark of 20 million ordinary read requests a month, using each vendor’s published rates.

Provider	Pricing Model	Free Tier	Entry Price	Cost at 20M requests
⭐ NOWNodes	Request quota	100K req, 1 month	€20 (Pro, 1M)	€200 (Business, 30M)
QuickNode	API credits	10M credits, 1 month	34–49 (Build)	Depends on credit weights
Alchemy	Compute units	30M CU/month	$0.525 per 1M CU	About $284
Chainstack	Request units	3M req/month	$49 (Growth, 20M)	$49
dRPC	Per request	210M CU/month	$6 per 1M requests	$120

Rates, supported chains, and limits change frequently. Verify current terms on each provider’s official site before committing.

Total Cost Analysis

NOWNodes. The €200 assumes the Business tier, whose 30 million quota covers the benchmark. On Pro Plus, the 10 million overage would add €500 at €5 per 100,000 requests, so quota size matters more than the base fee.

QuickNode. Credit weights vary by method, so 20 million requests may fit inside Build’s 80 million credits or overrun them. The $34 rate requires annual billing, while monthly billing costs $49.

Alchemy. The estimate uses Alchemy’s own default of 27 compute units per request. The 30 million free units cover only about 1.1 million requests at that weight, and log-heavy traffic costs more.

Chainstack. Growth’s 20 million request units match the benchmark exactly. Any extra call costs $15 per million, and archive calls count double.

dRPC. Flat $6 pricing keeps the arithmetic simple, and archive calls cost the same as full-node calls. At this volume the bill is roughly 2.4 times Chainstack’s Growth plan.

Do You Need HyperEVM Only, or HyperEVM Plus HyperCore?

Most teams need only one layer, and that choice splits these five providers into two groups. Every vendor here is a multi-chain generalist, so layer coverage separates them better than chain count does.

HyperEVM-focused providers. NOWNodes, Alchemy, and dRPC center on standard Ethereum JSON-RPC for contracts and wallets. NOWNodes also exposes 26 Hyperliquid info methods for reading account state. None of the three lists HyperCore streaming.

Providers that also serve HyperCore. QuickNode documents HyperCore info, WebSocket, and gRPC streams, and Chainstack states HyperEVM and HyperCore access. These suit order-book consumers and market-data pipelines, usually at a higher entry cost.

Which Hyperliquid RPC Provider Should You Choose?

NOWNodes suits teams that run Hyperliquid next to many other chains and want no fixed RPS ceiling on paid plans, provided they accept a mainnet-only listing and a 100,000-request trial. QuickNode suits systems that consume HyperCore streams and need audit reports, at a cost of at least the Build tier. Alchemy fits HyperEVM developers who value webhooks, data APIs, and a 30 million CU free allowance. Chainstack fits teams that want both layers, a permanent free plan, and a path to dedicated nodes. dRPC fits cost-led teams that need a low-price second endpoint behind a primary one.