Best Ethereum RPC providers for production workloads in 2026: pricing, uptime and free plans compared

Ethereum’s continued growth in 2026 has led to greater demand for high-performance Ethereum RPC providers. The network still handles roughly 2 million transactions per day, spanning smart contracts, rollups, trading bots, and stablecoin transfers — so choosing the right Ethereum RPC provider is now a production-critical decision.

Quick verdict: best Ethereum RPC providers in 2026

Chainstack is our top pick for production Ethereum workloads, based on request-based pricing, published RPS tiers up to ~600 RPS on Business, SOC 2 Type II and ISO 27001 certification, and the lowest median and tail latency in Chainstack’s own October 2026 benchmark of the three listed providers it measured (Chainstack, Alchemy and Quicknode). Chainstack publishes this page, runs that benchmark and is one of the 8 listed providers, so see how we compared and the benchmark limits before relying on the ranking. The best fit by need:

• Best for production workloads that need balanced pricing, uptime and compliance: Chainstack (request-based billing, Unlimited Node add-on, SOC 2 Type II + ISO 27001).
• Best for multi-chain teams that want archive and Blockbook access: NOWNodes (120+ chains under one API key).
• Best for flat per-response billing and dedicated archive clusters: Dwellir (1 credit per response, no compute-unit weighting).
• Best for teams pairing RPC with managed indexers: OnFinality (managed SubQuery indexers, Dedicated Nodes).
• Best for pay-as-you-go testing and smaller projects: Ankr (200M credits per month on the free tier).
• Best for Ethereum-first projects that use MetaMask: Infura (built-in MetaMask integration).
• Best for developer tooling: Alchemy (Webhooks, Transfers API, Debug/Trace APIs; fastest US P95 in the October 2026 benchmark).
• Best for high-throughput apps that need Streams and Webhooks: Quicknode (up to 500 RPS on Business).

Each project optimizes for different priorities – some need low latency, others cost-effectiveness, or reliability. In this comparison, we look at the best Ethereum RPC providers for web3 in 2026 across key metrics like speed, uptime, pricing, and security compliance.

We focus on two critical Ethereum use cases – stablecoin payments and enterprise applications – to see how each provider measures up.

Ethereum RPC providers compared: pricing, free plan, and uptime (2026)

What separates one RPC provider from another usually comes down to the same few metrics: pricing, latency, uptime, and scalability options available. Here’s how the leading Ethereum options measure up:

Rank	Provider	Best for	Free plan	Paid plans & pricing	Latency & uptime	Developer Experience
1	Chainstack	Production workloads that need balanced pricing, uptime, and compliance	3M requests/mo, 25 RPS	Starting from 250 RPS and scaling beyond 600 RPS on custom Enterprise plans, with an Unlimited Node add-on for unmetered requests at a flat monthly rate	99.99%+ uptime (status.chainstack.com, Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026). Oct 2026 benchmark (Chainstack-run): P50 36 ms, P95 78 ms, P99 117 ms, 99.98% probe success rate. Ethereum latency benchmark	Docs, dashboard metrics, testnet faucets (Sepolia/Hoodi), archive, Access rules
2	NOWNodes	Multi-chain teams that want one API key with archive and Blockbook access	Start plan: 100K requests/mo, 15 RPS, no WebSocket	Pro 1M to Enterprise 100M requests/month; overage from €5 per 100K (Pro Plus) down to €0.50 (Enterprise); no predefined RPS cap on paid plans (provider-stated); dedicated nodes on custom plans	99.95% stated uptime on shared nodes, 99.99% on dedicated; first support response within 3 minutes. Not in the Oct 2026 benchmark	Mainnet and Sepolia with RPC, WebSocket, Blockbook, and archive endpoints, debug/trace namespaces, 120+ chains under one API key
3	Dwellir	Flat per-response billing and dedicated archive clusters	20 responses/s, 100k responses/day	Developer $49/mo (25M responses, 100/s); Growth $299/mo (150M, 500/s); Scale $999/mo (500M, 5,000/s); Unlimited Nodes from $119/mo	99.99% uptime SLA stated on its marketing pages (contract terms not read); public status page. Not in the Oct 2026 benchmark	1 credit per response (no CU weighting), trace/debug on paid plans, HTTP + WebSocket, dedicated clusters from $2,400/mo (Ethereum archive cluster price not verified), crypto payments
4	OnFinality	Teams pairing RPC with managed SubQuery indexers	400K RU/day, 40 RU/sec, archive access	Growth $49/mo (20M RU, 200 RU/sec); Accelerate $249/mo (100M RU, 500 RU/sec); Ultimate custom	No published SLA on free/public tiers. Not in the Oct 2026 benchmark	Dedicated Nodes, trace/debug on paid tiers, WebSocket subscriptions, managed SubQuery indexers
5	Ankr	Pay-as-you-go testing and smaller projects	200M credits/month, ~30 RPS	Pay-as-you-go ($10 per 100M credits), up to about 1,500 RPS per endpoint on Premium	GEO-distributed network, Enterprise: contractual SLA up to 99.99% (99.9% standard), none on self-serve. Not in the Oct 2026 benchmark	Simple dashboard, multi-chain support, HTTPS & WebSocket, archive available
6	Infura	Ethereum-first projects that use MetaMask	3M credits/day (~depends on method), ~500 credits/sec	$50–$225/month for higher limits	Multi-region network, ~99.9% uptime. Not in the Oct 2026 benchmark	Ethereum-first with MetaMask integration, archive, debug/trace on paid tiers
7	Alchemy	Developer tooling: Webhooks, Transfers API, Debug/Trace	30M CUs/mo; ~15 RPS (300 CU/s)	Pay-as-you-go or monthly; higher CU throughput on paid plans	Uptime SLA on Enterprise only. Oct 2026 benchmark (Chainstack-run): P50 38 ms, P95 103 ms, P99 188 ms, 99.98% probe success rate. Ethereum latency benchmark	Webhooks (Notify), Transfers API, Gas Manager, Debug/Trace APIs, SDK support
8	Quicknode	High-throughput apps that need Streams and Webhooks	No free tier; 10M credits trial only	50–500 RPS; credit-based (method-weighted) monthly plans	99.99% uptime stated by the provider on paid tiers. Oct 2026 benchmark (Chainstack-run): P50 40 ms, P95 92 ms, P99 142 ms, 99.98% probe success rate. Ethereum latency benchmark	Streams, Webhooks, team dashboards, multi-region setup

Official pricing pages: Chainstack, NOWNodes, Dwellir, OnFinality, Ankr, Infura, Alchemy, Quicknode.

Key takeaways from the comparison:

• Chainstack offers a balanced feature set with transparent request-based pricing, 99.99%+ uptime (status.chainstack.com, Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026), and full SOC 2 Type II and ISO 27001 certification
• NOWNodes covers 120+ chains under one API key and exposes Ethereum RPC, WebSocket, Blockbook, and archive endpoints, with debug and trace APIs
• Dwellir bills one credit per response with no compute-unit weighting, and adds flat-rate Unlimited Nodes and dedicated archive clusters for teams that outgrow shared plans
• OnFinality offers a request-unit pricing model with Dedicated Nodes and managed SubQuery indexers on every paid tier
• Ankr offers flexible pay-as-you-go pricing with broad multi-chain support
• Infura remains reliable for Ethereum-first projects with deep MetaMask integration
• Alchemy provides the richest developer tooling including Webhooks, Transfers, and Debug/Trace APIs
• Quicknode offers globally distributed infrastructure and scales to 500 RPS on its Business tier

When comparing Ethereum RPC providers side by side, differences in pricing models, uptime guarantees, and throughput limits become especially clear at production scale.

Before diving into detailed provider reviews, let’s examine how each performs across critical Ethereum use cases:

How we compared Ethereum RPC providers

Disclosure: this comparison is published by Chainstack. Chainstack is one of the 8 providers listed and is ranked first, so read the ranking as a vendor perspective and check the figures against the official pricing pages linked in each provider section.

Criteria. Each provider is compared on the same attributes: pricing model, free tier, published rate limits (RPS), uptime and SLA evidence (platform uptime, probe success rate and contractual SLAs are kept separate), security certifications, and developer tooling. The order is editorial judgment based on those attributes for production workloads on Ethereum. It is not a benchmark result. The latency benchmark below covers only 3 of the 8 listed providers (Chainstack, Alchemy and Quicknode), so it informs the latency cells for those three and does not determine the order.

Sources. Prices, limits and uptime figures come from each provider’s public pricing, documentation and status pages. Chainstack’s platform uptime comes from status.chainstack.com (Mainnet component of each chain; testnet and devnet components are tracked separately), 90 days to October 29, 2026; other providers’ uptime figures are stated by the providers. Latency figures for Chainstack, Alchemy and Quicknode come from the Chainstack Labs Research Team’s October 2026 Ethereum benchmark. Where a provider publishes no Ethereum-specific latency figure and was not benchmarked, the provider section says so instead of estimating one. Plans change often, so confirm current prices on the linked official pricing pages before you buy.

Not covered. Other Ethereum RPC providers may also exist beyond these 8. They are not part of this comparison. dRPC appears in the benchmark table below because the report measured it, but it is not one of the 8 listed providers.

Ethereum RPC latency benchmark (October 2026): Chainstack, Alchemy and Quicknode

Disclosure: Chainstack operates the monitoring service, the public Grafana dashboard and the scoring formula behind this benchmark, and is one of the four providers measured. The figures below come from the Ethereum RPC provider benchmark (Research Note No. 050) by the Chainstack Labs Research Team, and the live data is on Chainstack Compare.

Method in brief: a 7-day window ending October 8, 2026; probes every 1 to 3 minutes from three cloud regions (Germany, US and Singapore) using a Python service on Vercel serverless functions; each provider on its first paid plan; a request counts as failed if it takes longer than 55 seconds. The score is response time divided by success rate cubed per region (lower is better), averaged across the three regions. The collection code is open source (compare-dashboard-functions).

Provider	Listed in this article	Score (lower is better)	Probe success rate	P50 (ms)	P95 (ms)	P99 (ms)	Germany P95 (ms)	US P95 (ms)	Singapore P95 (ms)
Chainstack	Yes	0.079	99.98%	36	78	117	47	143	42
Quicknode	Yes	0.097	99.98%	40	92	142	40	98	138
Alchemy	Yes	0.114	99.98%	38	103	188	64	80	164
dRPC	No (measured, not listed here)	0.670	99.96%	56	641	1,950	153	875	895

Source: Chainstack Labs Research Team, Ethereum RPC provider benchmark, October 2026 (7-day window ending October 8, 2026).

What the benchmark shows:

• Among the three listed providers it measured, Chainstack has the lowest P50 (36 ms), P95 (78 ms) and P99 (117 ms) and the best overall score (0.079). The report puts second-place Quicknode’s score 23% higher.
• Chainstack is not fastest everywhere. In the US, Alchemy (80 ms P95) and Quicknode (98 ms) beat Chainstack (143 ms). In Germany, Quicknode (40 ms) beats Chainstack (47 ms). Chainstack is lowest in Singapore (42 ms), and its best-to-worst region ratio is 3.4x (42 ms in Singapore against 143 ms in the US).
• Chainstack, Quicknode and Alchemy tie on probe success rate at 99.98% (the dashboard labels this “availability”; dRPC is 99.96%). This is the share of benchmark probe requests that succeeded over 7 days. It is not platform uptime and not an SLA. Chainstack’s platform uptime is a separate figure: 99.99%+ for the Ethereum Mainnet component over the 90 days to October 29, 2026 on status.chainstack.com (shown by the status page, not a guarantee). The status page covers Chainstack only, so this page makes no uptime comparison with other providers.
• dRPC, which is not one of the 8 listed providers, is degraded in the US (875 ms P95) and Singapore (895 ms P95).

What this benchmark does not show: NOWNodes, Dwellir, OnFinality, Ankr and Infura were not measured, so this page does not claim Chainstack is the fastest of the 8 listed providers. The report covers a single 7-day window, each provider on its first paid plan only (free and higher plans can behave differently), and synthetic probes from three cloud regions rather than end-user latency. It does not measure throughput under load, cost or correctness, and because Chainstack chose the scoring formula, a different formula could reorder close pairs.

To check latency from the region where your application runs, time one eth_blockNumber call against each endpoint. Repeat it and compare the median and the 95th percentile. See how to measure RPC latency (P50, P95, P99) for the full method.

curl -s -o /dev/null -w "%{time_total} s" -X POST https://YOUR_ENDPOINT -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","method":"eth_blockNumber","params":[],"id":1}'

Choose Ethereum RPC providers by use case

Stablecoin & payments infrastructure

Stablecoins have become the backbone of on-chain payments, now making up ~30% of all crypto transaction volume (over $4 trillion in 2025). Payment applications require Ethereum RPC service with consistent throughput, near-zero downtime, and predictable costs. Even minor RPC outages or high latency can disrupt thousands of stablecoin transfers.

For teams building payment infrastructure with strict requirements around gas cost predictability and settlement finality, purpose-built chains like Tempo offer an alternative optimized specifically for stablecoin operations. Tempo delivers sub-second finality with deterministic execution costs designed for enterprise payment rails.

Top providers address this with robust multi-region infrastructure that automatically fails over if any node goes down. Chainstack’s network delivered 99.99%+ uptime (status.chainstack.com, Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026), with multi-cloud architecture that reroutes traffic instantly if a region fails. Such redundancy ensures stablecoin transactions aren’t halted by regional outages.

Enterprise Ethereum RPC providers

Enterprise Ethereum deployments demand strong performance guarantees. Corporate and institutional users often have high-throughput workloads and strict SLA requirements.

Enterprise RPC essentials:

• High throughput: Chainstack offers high-throughput plans with custom RPS configurations on Ethereum and achieves 99.99%+ platform uptime (status.chainstack.com, Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026) via global load balancing
• SLA guarantees: Quicknode processes hundreds of billions of API calls a month, states 99.99% uptime, and offers a premium SLA for enterprise clients
• Low-latency across regions: Multi-region data centers and intelligent routing keep response times uniform worldwide

NFT & indexing infrastructure

NFT marketplaces, indexers, and analytics platforms need to replay full historical state — every mint, transfer, and contract event since genesis — rather than just the latest block. That means archive access, debug/trace support, and enough throughput to backfill large datasets without throttling.

NFT & indexing essentials:

• Full archive history: Chainstack Archive nodes serve complete historical blocks from genesis to the latest one, including debug and trace queries
• Bulk backfill throughput: Custom RPS configurations let indexers pull large historical ranges without hitting plan limits
• Cost-efficient scaling: The Unlimited Node add-on caps costs with flat monthly pricing, which matters when reindexing large collections

SOC 2–compliant Ethereum RPC providers

Security and compliance are paramount for enterprises integrating with Ethereum. SOC 2 certification has become a key benchmark of a provider’s security controls and operational integrity.

Security leaders:

• Chainstack holds SOC 2 Type II and ISO 27001 certification; Quicknode also holds SOC 2 Type II and ISO 27001, reflecting rigorous audits of security, availability, and confidentiality practices

Note: Not all popular RPC services have completed SOC 2 certification. Providers without published attestation may still meet security requirements internally, but they can’t offer the audit-backed evidence that regulated institutions typically require during vendor review.

Ethereum RPC rate limits and free tiers compared

Rate limits and billing units differ between providers, so the same number of calls can cost different amounts. This table summarizes the figures from the provider sections below.

Provider	Free tier	Free-tier rate limit	Highest published paid rate limit	Billing unit
Chainstack	3M requests/month	~25 RPS	~600 RPS on Business; custom on Enterprise	Requests
NOWNodes	100,000 requests/month (Start plan)	15 RPS, no WebSocket	No predefined RPS cap on paid plans (provider-stated)	Requests
Dwellir	100,000 responses/day	20 responses/sec	5,000 responses/sec on Scale (10,000 burst)	Responses
OnFinality	400,000 request units/day	40 RU/sec (public endpoint: 5 RPS)	500 RU/sec on Accelerate; custom on Ultimate	Request units
Ankr	200M credits/month	~30 RPS	Up to about 1,500 RPS on Premium; Enterprise RPS not verified	Credits (method-weighted)
Infura	3M credits/day	500 credits/sec	40K credits/sec on Team; custom on Enterprise	Credits (method-weighted)
Alchemy	30M CUs/month	300 CU/sec (about 15 RPS)	From 300 RPS on pay-as-you-go; custom on Enterprise	Compute units
Quicknode	No free tier; 10M-credit trial	15 RPS (trial)	500 RPS on Business; custom on Enterprise	Credits (method-weighted)

Chainstack alternatives for Ethereum RPC

If Chainstack is not the right fit, these are the head-to-head differences between Chainstack and each alternative on Ethereum:

• Chainstack vs NOWNodes: NOWNodes covers 120+ chains under one API key and adds Blockbook, but publishes no SOC 2 or ISO 27001 attestation and caps its free Start plan at 15 RPS without WebSocket; Chainstack holds both certifications and offers 3M requests/month at ~25 RPS free.
• Chainstack vs Dwellir: both offer flat-rate unlimited options (Unlimited Node add-on from $149/month; Dwellir Unlimited Nodes from $119/month); Dwellir bills one credit per response and offers dedicated archive clusters, but its public pricing pages show no SOC 2 or ISO 27001 attestation.
• Chainstack vs OnFinality: OnFinality adds managed SubQuery indexers and Dedicated Nodes on every paid tier, but publishes no SLA on free/public tiers and caps its public endpoint at 5 RPS; Chainstack publishes fixed RPS tiers.
• Chainstack vs Ankr: Ankr has a larger free credit allowance (200M credits/month) and pay-as-you-go pricing; its SLA applies to Enterprise only, while Chainstack publishes fixed RPS tiers.
• Chainstack vs Infura: Infura has built-in MetaMask integration and archive support, but credit-based pricing makes cost depend on the method mix and it publishes no Ethereum-specific latency metrics; Chainstack bills per request.
• Chainstack vs Alchemy: Alchemy adds Webhooks, the Transfers API and Debug/Trace APIs and had a faster US P95 (80 ms against 143 ms) in the October 2026 benchmark, but it bills in compute units, so cost depends on the call mix. Chainstack had the lower overall P50, P95 and P99.
• Chainstack vs Quicknode: Quicknode reaches 500 RPS on Business, adds Streams and Webhooks, and had faster P95 in Germany and the US in the October 2026 benchmark, but uses method-weighted credits and has no flat-rate unlimited option. Chainstack had the lower overall P50, P95 and P99.

In-depth Ethereum RPC providers analysis: pricing, free plan, and limits

Below, we take a closer look at how individual Ethereum RPC providers perform in real-world production environments, including latency consistency, scalability, and operational reliability.

1. Chainstack: best for production workloads that need balanced pricing, uptime and compliance

Verdict: RPS is fixed per tier, so ultra-low-latency apps may need Dedicated Nodes (Pro plan or higher); in Chainstack’s own October 2026 benchmark its US P95 (143 ms) was slower than Alchemy and Quicknode. Official pricing: Chainstack pricing.

Best for: Production workloads that need balanced pricing, uptime, and compliance.

Chainstack is a multi-chain infrastructure platform offering one of the most complete ways to connect to Ethereum without the overhead. By choosing Chainstack Ethereum RPC node, you get secure HTTP and WebSocket access to Ethereum Mainnet, Sepolia, and Hoodi, backed by 99.99%+ uptime and globally distributed routing for consistently low latency.

What makes it stand out is how much is available from the same console:

• Global Nodes for geo-balanced Ethereum RPC access.
• Dedicated Nodes for isolated performance and full control over node configuration.
• Archive nodes for deep historical reads, full historical blocks from the genesis to the latest one, debug and trace queries on the Ethereum blockchain.
• Access rules to manage allowed domains and IPs directly from the dashboard.
• Unlimited Node add-on for unlimited requests with flat monthly pricing, so you never have to worry about hitting plan limits.
• Testnet faucets for Sepolia and Hoodi are available.

All of this adds up to the Ethereum setup that covers nearly every use case. Whether you’re deploying contracts, building DeFi products, or streaming on-chain data, Chainstack gives you reliable performance, clear pricing, and infrastructure to scale with.

How much does Chainstack cost?	Developer — free, 3M requests/month (~25 RPS), $20 per extra million. Growth — $49/month, 20M requests (~250 RPS), $15 per extra million. Pro — $199/month, 80M requests (~400 RPS), $12.5 per extra million. Business — $499/month, 200M requests (~600 RPS), $10 per extra million. Enterprise — from $990/month, 400M requests (custom RPS), extra from $5 per million. Unlimited Node Add-on — flat monthly fee for unmetered requests starting from $149, priced by RPS tier. Dedicated Nodes — from $0.50 / hour per node plus storage ($0.01 per 20 GB/hour) for exclusive, high-performance isolated Ethereum node instances, with an effective per-request cost that falls as volume grows. While others price in shifting compute units, Chainstack keeps it simple with request-based billing and clear RPS tiers, so you always know what you’ll pay. What’s more, on Chainstack, you are able to pay in crypto.
Chainstack performance	Uptime: 99.99%+ platform uptime (status.chainstack.com, Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026), backed by a multi-cloud network that reroutes automatically if a region drops. Public status page is also available. Latency: Choose endpoints in the US, EU, or APAC to keep response times tight. In the October 2026 benchmark (Chainstack-run) Chainstack measured P50 36 ms, P95 78 ms and P99 117 ms overall, with a US P95 of 143 ms. Throughput: Plans come with enough RPS capacity to absorb traffic surges without forcing you to throttle the app. Stability: 99.98% probe success rate in the October 2026 Ethereum benchmark, tied with Quicknode and Alchemy; platform uptime was 99.99%+ on status.chainstack.com (Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026) (see the Ethereum latency benchmark). Monitoring: Built-in console metrics plus public performance/compare dashboards and a status page.
Chainstack pros	– High throughput at price point (~250 RPS on Growth, ~600 RPS on Business) – Predictable, request-based pricing with Unlimited Node add-on for flat, unmetered traffic within your chosen RPS tier – Built-in testnet faucets for Sepolia and Hoodi – 99.99%+ platform uptime (status.chainstack.com, Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026), SOC 2 Type II + ISO 27001, globally distributed routing for low latency One provider for Ethereum plus 70+ other chains
Chainstack cons	– Fixed RPS per tier; ultra-low-latency apps may need Dedicated Nodes – Free plan (3M calls, ~25 RPS) can be tight for bots or indexers – Dedicated Nodes require at least a Pro plan

🤖 You can also access Chainstack Ethereum RPC directly from Claude, Cursor, Codex, Gemini, or Windsurf using Chainstack MCP. Learn more about Chainstack MCP.

With Chainstack, you get the full stack: reliability, high RPS, global routing, and archive access, all with clear, request-based pricing. From testing to scaling, the setup doesn’t change, and the costs stay predictable, making it one of the most dependable ways to run Ethereum at any stage.

2. NOWNodes: best for multi-chain teams that want archive and Blockbook access

Verdict: No SOC 2 or ISO 27001 published, and the free Start plan is capped at 15 RPS with no WebSocket. Official pricing: NOWNodes pricing.

Best for: Multi-chain teams that want one API key with archive and Blockbook access.

NOWNodes is a multi-chain node provider covering 120+ blockchains under a single API key. For Ethereum it offers mainnet and Sepolia, with separate JSON-RPC, WebSocket, Blockbook, and archive endpoints, and its docs cover the debug and trace namespaces alongside the standard eth_* methods.

NOWNodes suits teams that already run several chains through one provider and want indexed address and transaction data from Blockbook without building it themselves, though it publishes no SOC 2 or ISO 27001 attestation.

How much does NOWNodes cost?	Plans are metered in monthly requests. The free Start plan includes 100,000 requests a month with one API key and a 15 RPS cap, and WebSocket is not part of it (NOWNodes’ own FAQ describes Start as a free month, so confirm it renews). Paid plans run from 1 million (Pro) to 100 million (Enterprise) requests a month, with overage of €5 per 100K on Pro Plus down to €0.50 on Enterprise. Dedicated nodes are available through custom plans.
NOWNodes performance	Uptime: 99.95% stated on its shared-node page, 99.99% on its dedicated-node page. Throughput: No predefined RPS limits on paid plans, scaling with cluster capacity (provider-stated); 15 RPS on Start. Support: First response within 3 minutes on every plan.
NOWNodes pros	– 120+ chains under one API key, – Archive, debug, and trace endpoints on Ethereum mainnet, – Blockbook indexed API alongside RPC
NOWNodes cons	– No SOC 2 or ISO 27001 published, – Free Start plan has a 15 RPS cap and no WebSocket, – Docs list different hosts for debug methods, so test the endpoint before relying on it

NOWNodes is a reasonable fit when one account has to cover many chains and you need archive and Blockbook access on Ethereum. Teams that need audit-backed controls should confirm certifications directly, since none are published.

3. Dwellir: best for flat per-response billing and dedicated archive clusters

Verdict: No SOC 2 or ISO 27001 attestation found on its public pricing pages, the free tier is capped at 100,000 responses per day, and dedicated clusters are listed from $2,400/month (the Ethereum archive cluster price was not verified). Official pricing: Dwellir pricing.

Best for: Flat per-response billing and dedicated archive clusters.

Dwellir is a Swedish RPC provider with a long Polkadot track record that now serves 150+ networks, Ethereum among them. For Ethereum it offers a managed API over HTTPS and WebSocket, with trace and debug methods on paid plans, plus dedicated archive clusters for teams that want isolated, unmetered access.

Shared plans bill one credit per response with no compute-unit weighting, so a heavy call such as eth_getLogs costs the same as eth_blockNumber. Plans scale by responses per month and responses per second, and you can pay in crypto.

How much does Dwellir cost?	Free — 20 responses/second, 100,000 responses per day. Developer — $49/month, 25M responses, 100 responses/second. Growth — $299/month, 150M responses, 500 responses/second, eth_getLogs over up to 10,000 blocks. Scale — $999/month, 500M responses, 5,000 responses/second (10,000 burst). Unlimited Nodes — from $119/month flat, 25 to 1,000 RPS. Dedicated archive cluster — clusters are listed from $2,400/month per chain, unmetered (the Ethereum archive cluster price was not verified). Pay-as-you-go is available above 1B responses per month.
Dwellir performance	Uptime: its marketing pages state a 99.99% uptime SLA (contract terms not read); a public status page covers the shared service. Latency: No Ethereum-specific latency figures published on its product pages. Throughput: Plan-based, up to 5,000 responses/second on Scale; no rate limits or request counting on a dedicated cluster. Monitoring: Usage dashboard; chat support on Developer, a dedicated account manager from Growth.
Dwellir pros	– 1 credit per response, no compute-unit weighting – Trace and debug APIs on paid plans, HTTP and WebSocket on every plan – Flat-rate Unlimited Nodes and dedicated archive clusters in the same account – Crypto payments accepted
Dwellir cons	– No SOC 2 or ISO 27001 attestation found on its public pricing and dedicated-cluster pages – Free tier is capped at 100,000 responses per day – Dedicated clusters are listed from $2,400/month (Ethereum archive cluster price not verified)

Dwellir is a good fit for teams that want predictable per-response billing, trace and debug without a premium tier, and a path from shared plans to a dedicated archive cluster. Ask for its compliance documents up front if your vendor review requires SOC 2 or ISO 27001.

4. OnFinality: best for teams pairing RPC with managed SubQuery indexers

Verdict: No SLA is published on free/public tiers and the public endpoint is capped at 5 RPS. Official pricing: OnFinality pricing.

Best for: Teams pairing RPC with managed SubQuery indexers.

OnFinality runs managed Ethereum infrastructure on a request-unit (RU) model across four tiers: Developer (free, 400K RU/day, 40 RU/sec, archive access, no trace API), Growth ($49/mo, 20M RU/month, 200 RU/sec, trace API included, $6 per million RU overage), Accelerate ($249/mo, 100M RU/month, 500 RU/sec, $3.75 per million RU overage), and Ultimate (custom limits and pricing, enterprise support). The public endpoint is capped at 5 RPS. Dedicated Ethereum Nodes are available on every paid tier for high-volume RPC, exchanges, indexers, archive-heavy analytics, and trace workloads.

Pros: Dedicated Nodes, archive plus trace/debug on paid tiers, WebSocket subscriptions, managed SubQuery indexers for teams building analytics on top of raw RPC. Cons: No SLA published on free/public tiers; paid-tier SLA terms are best confirmed with OnFinality before a production commitment.

5. Ankr: best for pay-as-you-go testing and smaller projects

Verdict: The contractual SLA applies to Enterprise only, there are no fixed monthly plans, and performance can vary across regions. Official pricing: Ankr pricing.

Best for: Pay-as-you-go testing and smaller projects.

Ankr provides Ethereum RPC through a globally distributed network of node operators, routing traffic across regions to keep latency low and reduce reliance on any single infrastructure provider. You can connect to Ethereum mainnet and testnets like Sepolia via HTTPS or WebSocket, with more than 30 regions in rotation and archive access available for historical queries and tracing

Ankr is a solid option for teams that prefer a pay-as-you-go model and want to tap into a decentralized setup, though Ankr skips some of the advanced tooling and analytics you’d find in larger, more centralized providers.

How much does Ankr cost?	Ankr uses an API credits model where credit cost varies by method (heavier calls cost more). Pricing is pegged at $0.10 = 1M API credits (PAYG), and there’s a freemium tier. It’s enough for testing, but high-traffic apps will hit the ceiling quickly. Their paid plans use a pay-as-you-go model: Premium — $10 per 100M credits (~500K requests), with up to 1,500 RPS. Enterprise — deal plans from $500/month with 20% extra credits, dedicated endpoints and priority support (Enterprise RPS ceiling not verified).
Ankr performance	Uptime: Contractual SLA on Enterprise only (99.9% standard, up to 99.99% negotiable); none on self-serve. Latency: Region-routed. Throughput: 30 RPS on free; up to about 1.5K RPS per endpoint on Premium (higher tiers not verified). Reliability: Distributed design reduces single-point failures but outages can still occur. Monitoring: The console shows usage and request stats, with more advanced visibility once you move to a paid.
Ankr pros	– Freemium with 200M credits/month, – Supports HTTP, WebSocket, and archive queries, – Multi-chain support
Ankr cons	– Advanced visibility/support gated to paid, – No fixed monthly plans, – Fewer built-in analytics

If you don’t mind variable performance across regions, Ankr could be a way to go when testing or for the smaller projects. Yet, if you need steady throughput and predictable billing, platforms with fixed RPS tiers and dedicated routing tend to offer a more stable path for production workloads.

6. Infura: best for Ethereum-first projects that use MetaMask

Verdict: Credit-based cost depends on the method mix, IPFS access requires pre-approval, and no Ethereum-specific latency metrics are published. Official pricing: Infura pricing.

Best for: Ethereum-first projects that use MetaMask.

Infura provides Ethereum RPC access through infrastructure operated by ConsenSys, with direct integration into MetaMask for wallet connections. You get Ethereum mainnet and testnets like Sepolia via HTTPS and WebSocket, including archive data for historical queries. The service focuses on Ethereum-first support, with add-ons for Layer 2s (L2s) such as Polygon and Optimism.

How much does Infura cost?	Infura uses a credit-based model where each request consumes credits based on method complexity. This structure works for Ethereum-focused apps but demands monitoring to avoid caps on high-volume days. Core (Free) — 3M credits/day, 500 credits/sec. Developer — $50/month, 15M credits/day, 4K credits/sec. Team — $225/month, 75M credits/day, 40K credits/sec. Enterprise — custom.
Infura performance	Uptime: Targets about 99.9%; the pricing page lists no SLA figure. Latency: No official ms target published. Throughput: Credit-per-second caps by plan, Effective RPS depends on method credit cost (see examples above). Reliability: Maintains performance during spikes. Monitoring: Dashboard with usage analytics and request visibility (24h on Free; up to 30 days on paid).
Infura pros	– Built-in MetaMask integration, – Broad L2 coverage, – Archive support and Debug/Trace on eligible plans
Infura cons	– IPFS API/Gateway access requires pre-approval, not open for all users, – Cost depends on method mix, – No public Ethereum-specific latency metrics

Infura is a good choice for Ethereum-focused projects, especially those that rely on MetaMask. The main drawback is its credit-based structure (and IPFS access gating), which introduces limits and variable costs as usage grows. For teams that are comfortable with those trade-offs, it would be a good pick.

7. Alchemy: best for developer tooling: Webhooks, Transfers API, Debug/Trace

Verdict: Compute-unit pricing depends on the call mix and very high RPS needs an Enterprise plan; in Chainstack’s October 2026 benchmark Alchemy had the lowest US P95 (80 ms) but a 103 ms overall P95. Official pricing: Alchemy pricing.

Best for: Developer tooling: Webhooks, Transfers API, Debug/Trace.

Alchemy offers Ethereum RPC through its Supernode, giving developers access to mainnet and testnets over HTTP and WebSocket, along with extras like Webhooks (Notify), the Transfers API, Gas Manager, and Debug/Trace APIs.

On top, Alchemy supports dozens of networks, giving access to historical blockchain data, real-time event streaming, and built-in MEV protection on some chains. Backed by global infrastructure, Alchemy suits teams looking for additional APIs and monitoring tools.

How much does Alchemy cost?	Alchemy measures usage in compute units (CUs) rather than raw requests, which can make costs less predictable for teams sending large volumes of latency-sensitive calls compared to flat, request-based models. Alchemy starts with a free tier offering 30M CUs per month, roughly 1.1M typical calls at about 15 RPS. Paid plans scale by total CU usage: Pay-as-you-go — $0.525 per million CUs (from 300 RPS). Enterprise — custom pricing with lower per-unit costs at high volumes.
Alchemy performance	Uptime: Reports 99.9% historical uptime, but an uptime SLA is only guaranteed on Enterprise plans. Latency: Region-based routing; Alchemy does not publish official Ethereum-specific latency numbers. In the October 2026 benchmark (Chainstack-run) Alchemy measured P50 38 ms, P95 103 ms and P99 188 ms overall, with the lowest US P95 (80 ms) of the providers measured. Throughput: Begins at about 15 RPS (300 CU/s) on the free tier, from 300 RPS on pay-as-you-go, with more capacity bought in 5,000 CU/s blocks. Reliability: The Supernode setup helps maintain data accuracy during heavy traffic. Monitoring: You can check usage, errors, and request patterns through the dashboard in real time.
Alchemy pros	– Covers Ethereum plus 100+ chains, – Built-in MEV protection on supported networks, – Rich tooling
Alchemy cons	– CU pricing isn’t intuitive, – Alchemy publishes no Ethereum-specific latency benchmarks of its own (Chainstack’s October 2026 benchmark measured it), – Throughput caps by plan, very high RPS needs enterprise plan

In Chainstack’s October 2026 benchmark, Alchemy had the fastest US P95 of the providers measured (80 ms) but a slower Singapore P95 (164 ms). Alchemy’s compute-based pricing gives you flexibility in how you use the service, but it also means your bill depends on call mix, not just volume, which can make it harder to model at scale. If you’re building with features like Webhooks or Debug/Trace APIs, Alchemy is a fine option, but less ideal if you want predictable scaling.

8. Quicknode: best for high-throughput apps that need Streams and Webhooks

Verdict: Method-weighted credits, no free tier (10M-credit trial only) and no flat-rate unlimited option; its overall P95 was 92 ms in Chainstack’s October 2026 benchmark. Official pricing: Quicknode pricing.

Best for: High-throughput apps that need Streams and Webhooks.

Quicknode runs a multi-region infrastructure for Ethereum RPC, letting you connect to mainnet and testnets like Sepolia over HTTP or WebSocket. Archive support is available for teams that need to query deep historical data.

Quicknode also supports dozens of other chains and offers add-ons such as Webhooks and Streams if you want event-based triggers or real-time blockchain feeds. Quicknode is popular with high-traffic teams, but reaching its highest RPS means moving to the higher-priced tiers.

How much does Quicknode cost?	Build — $49/month, 80M credits plus $0.62 per extra million, up to 50 RPS. Accelerate — $249/month, 450M credits plus $0.56 per extra million, up to 125 RPS. Scale — $499/month, 950M credits plus $0.53 per extra million, up to 250 RPS. Business — $999/month, 2B credits plus $0.50 per extra million, up to 500 RPS. Enterprise — custom terms for higher volumes and dedicated support. Unused credits do not roll over, so size the plan to your monthly peak.
Quicknode performance	Uptime: States 99.99% uptime across supported services on its security page; premium SLA on Enterprise. Latency: Region-routed endpoints; no official ms target published, so benchmark in your region. In the October 2026 benchmark (Chainstack-run) Quicknode measured P50 40 ms, P95 92 ms and P99 142 ms overall, with the lowest Germany P95 (40 ms). Throughput: The trial is capped at 15 RPS and Build starts at 50 RPS; advanced tiers scale well past 500 RPS. Reliability: Requests are rerouted automatically. Monitoring: The dashboard gives real-time visibility into request performance and errors.
Quicknode pros	– Scales up to 500+ RPS on higher plans, – Archive data available (by chain/plan), – Low latency from multi-region routing
Quicknode cons	– Free trial caps at 15 RPS with only community support, – Method-weighted, credit-based billing makes costs less predictable, – No flat-rate unlimited option for heavy workloads

Quicknode offers global routing and developer tooling, which makes it a good option for Ethereum apps. The main limitation is its credit-based (method-weighted) pricing. Because different methods burn credits at different rates, long-term cost planning takes more effort, especially if your workload isn’t static.

Conclusion

Choosing the right Ethereum RPC providers in 2026 depends on your workload profile, compliance requirements, and tolerance for variable pricing models. The 2026 Ethereum RPC landscape is more competitive and mature than ever. In our assessment, Chainstack leads as the top all-around provider, combining enterprise-grade performance and 99.99% uptime via multi-cloud routing, transparent request-based pricing (about $2.45–$2.50 per million requests on the paid plans, with unlimited options), and full SOC 2 Type II and ISO 27001 compliance. Chainstack serves stablecoin payments and enterprise dApps, with 70+ chains supported from one platform.​

NOWNodes covers 120+ chains under one API key with archive, debug, and trace endpoints on Ethereum, though it publishes no SOC 2 or ISO 27001 attestation. Dwellir bills one credit per response with no compute-unit weighting and adds flat-rate Unlimited Nodes and dedicated archive clusters for teams that want isolation, though it publishes no SOC 2 or ISO 27001 attestation. OnFinality runs a request-unit pricing model with Dedicated Nodes and managed SubQuery indexers on every paid tier. Quicknode scales to 500 RPS on its Business tier for latency-critical apps, with dual SOC 2/ISO certifications. Alchemy offers rich developer tooling (Webhooks, Transfers, Debug/Trace APIs, generous free tier). Ankr provides flexible pay-as-you-go pricing with broad multi-chain support. Infura remains Ethereum’s trusted veteran with MetaMask integration.

Developers now have several production-grade options. The right one depends on workload, compliance needs and pricing model; check the benchmark limits before treating any latency figure as a ranking of all 8 providers.

FAQ

Not only do you get the option to choose between Global Nodes for shared, geo-balanced access or Dedicated Nodes for isolated performance and full control, but every deployment comes production-ready. Both options include Bolt fast sync for same-day readiness and GraphQL support.

You also get built-in security through Access rules, plus testnet faucets and MEV protection enabled by default. Everything required to connect, build, and ship on Ethereum is available in one place.

When evaluating Ethereum RPC node providers, it’s important to consider not only raw performance, but also long-term reliability, pricing transparency, and support for production workloads.

Reliable Ethereum RPC infrastructure

Getting started with Ethereum on Chainstack is fast and straightforward — deploy a reliable node in seconds through an intuitive console, no hardware or complex setup required.

With 99.99%+ platform uptime (status.chainstack.com, Ethereum Mainnet, 99.99%+ over the 90 days to October 29, 2026), 24/7 priority support on Enterprise and low-latency global endpoints, Chainstack ensures seamless RPC access for building and scaling DeFi, analytics, and trading applications.

Additional resources

For more on building with Ethereum RPC, explore these resources:

• Ethereum API reference on Chainstack — JSON-RPC methods, quickstart, and endpoint setup
• Ethereum tooling guide — wallets, SDKs, and libraries that work with Chainstack Ethereum nodes
• Chainstack Ethereum RPC — the product page for deploying Ethereum mainnet and testnet nodes
• Ethereum JSON-RPC specification — the official method-by-method spec on ethereum.org
• Ethereum developer docs — ethereum.org’s official documentation for building on Ethereum

Changelog

We review this comparison regularly so that the information stays accurate, clear, and easy to verify. This page was reviewed to improve how the comparison is structured and sourced.

• October 29, 2026: Added a quick verdict, a Rank column, a methodology and disclosure section, an Ethereum latency benchmark section (October 2026, covering Chainstack, Alchemy and Quicknode) with matching latency figures in the comparison table, a rate limits and free tiers comparison, alternatives guidance, one-line verdicts and official pricing links for each provider, corrected vendor pricing, limits and certification details, and wording clarifications.
• January 6, 2026: Original publication.