Black Friday ~40% off Growth and Pro annual plans See offers
  • Agents
  • Pricing

Chainstack vs Alchemy RPC benchmark: 7 networks, November 2026

Created Nov 16, 2026 Updated Nov 16, 2026
Chainstack Vs Alchemy logo

Data source: 27 “Chainstack vs. Alchemy” reports by CompareNodes.com, measured from Nov 4, 2026 to Nov 9, 2026 (13 batches over 6 days) from 30 global locations, with North America, Europe and Asia sub-sets.

📢 Disclosure. Chainstack publishes this article and is one of the two providers measured. The probes, charts and win counts come from the CompareNodes.com reports; this article summarizes them and adds no measurements of its own. Both providers were tested on paid accounts. Where Alchemy leads, the article says so.

⏭ Skip ahead: the Global results table has every network in one view.

Abstract

Chainstack and Alchemy were measured head to head on seven networks for six days in November 2026. On the Global 30-location cohort, Chainstack won more JSON-RPC p95 comparisons than Alchemy on five of six EVM networks (all except Ethereum), and led WebSocket newHeads on three of six (Base, HyperEVM, Robinhood). Alchemy led Ethereum JSON-RPC, newHeads on Ethereum, Arbitrum and BNB, and Sui gRPC checkpoints in the Global cohort by 2.9 points.

Key findings

  • JSON-RPC, Global: Chainstack won more regional p95 comparisons on five of six EVM networks. Arbitrum is the widest margin, 76.3% (2,084) vs 23.7% (646). Ethereum is the exception: Alchemy won 53.2% (1,452) vs 46.8% (1,278).
  • Asia is where Chainstack trails on JSON-RPC. Chainstack led Asia only on Arbitrum (71.1%) and Base (67.6%). Alchemy led Asia on Ethereum (75.2%), BNB Chain (64.6%), HyperEVM (61.8%) and Robinhood (69.9%).
  • Ethereum pending transactions: Chainstack won 82.1% (120,836) of newPendingTransactions first-arrival comparisons Global and led in every region; Alchemy won 17.8% (26,254).
  • Ethereum new blocks: Alchemy won 97.5% (4,215) of newHeads first-arrival comparisons Global; Chainstack won 2.3% (101). Alchemy also led newHeads on Arbitrum (57.8%) and BNB Chain (73.2%) Global.
  • Sui gRPC checkpoints: Alchemy 51.4% (28,884) vs Chainstack 48.5% (27,284) Global, a 2.9-point gap. Chainstack led North America (90.4%) and Europe (85.5%); Alchemy led Asia (94.5%).

Chainstack vs Alchemy RPC provider comparison: Global results by network

Global cohort: 30 AWS locations. JSON-RPC is the share of region-method p95 comparisons won (lower p95 wins). WebSocket and gRPC are the share of first-arrival comparisons won (higher is better). Each cell shows Chainstack / Alchemy, and bold marks the higher value.

NetworkJSON-RPC (7 methods)WSS newHeadsLeads
Ethereum46.8% / 53.2%2.3% / 97.5%Alchemy: JSON-RPC, newHeads
Arbitrum76.3% / 23.7%40.6% / 57.8%Chainstack: JSON-RPC
Alchemy: newHeads
Base66.6% / 33.4%85.3% / 13.2%Chainstack: JSON-RPC, newHeads
BNB Chain50.8% / 49.2%26.2% / 73.2%Chainstack: JSON-RPC (46 comparisons ahead)
Alchemy: newHeads
HyperEVM52.4% / 47.6%71.5% / 27.9%Chainstack: JSON-RPC, newHeads
Robinhood Chain54.1% / 45.9%76.4% / 23.3%Chainstack: JSON-RPC, newHeads

Two measures cover a single network each:

MeasureChainstackAlchemyLeads
Ethereum WSS newPendingTransactions82.1%17.8%Chainstack
Sui gRPC checkpoints48.5%51.4%Alchemy
Chainstack vs Alchemy Global results by network, Nov 4–9, 2026: Chainstack won 76.3% of Arbitrum JSON-RPC comparisons and 85.3% of Base newHeads comparisons; Alchemy won 53.2% of Ethereum JSON-RPC and 97.5% of Ethereum newHeads comparisons.

Percentages exclude ties, which stay in the denominator only, so a pair can sum to just under 100%. Chainstack’s lead on BNB Chain JSON-RPC is 1,388 vs 1,342 comparisons, a 1.6-point margin; the reports apply no significance test.

How the Chainstack vs Alchemy benchmark works

CompareNodes.com ran one AWS Lambda probe per endpoint per location, each with its own execution environment. Every Chainstack vs Alchemy run selects the two endpoints and measures them in the same locations at the same time. The test scenarios are described on CompareNodes’ RPC Inspector Pro page.

  • JSON-RPC over HTTPS. Seven requests in sequence: eth_getBlockByNumber, eth_getLogs by block hash, eth_getLogs by a 50-block range, eth_getBalance, eth_getTransactionReceipt, debug_traceBlockByNumber (callTracer, where supported) and eth_call (Multicall3 token balances). Target: 15 iterations in a 35-second loop, 1,100 ms between request starts, 4,000 ms timeout. Request round-trip time excludes pacing and earlier calls; setup and warm-up are excluded.
  • Scoring JSON-RPC. Lower p95 round-trip time wins each method-region comparison. Both providers need successful samples. More regional wins decides the run.
  • WebSocket. eth_subscribe to newHeads for 135 seconds and to newPendingTransactions for 35 seconds. The provider that delivers a matched block hash or transaction hash first, in the same region, wins that comparison. Unmatched events and ties earn neither provider a win.
  • Sui. gRPC checkpoint subscription, compared by first arrival.
  • Cohorts. Global is 30 AWS locations. North America has 6, Europe 8 and Asia 11. Each is measured as its own cohort.
  • Window. Nov 4, 2026 to Nov 9, 2026: 13 batches over six consecutive days. Ethereum has 39 runs (JSON-RPC, newHeads and newPendingTransactions, 13 each); the other EVM networks have 26; Sui has 13.

Limits

  • Six days, one window. Two windows are not a trend, and results can change between measurements.
  • Win counts, not latency distributions. The headline numbers say how often a provider was faster, not by how much. A narrow win and a wide win count the same. No significance test or minimum winning margin is applied.
  • Paid accounts, plan not stated. The reports say both providers were tested on paid accounts and do not name the plan. Other plans or dedicated nodes may behave differently.
  • Probe location is not user location. AWS Lambda probes measure round-trip time from cloud regions, not from end users, and the reports do not measure response correctness beyond recorded problems.
  • Global p95 can hide regional strength. In several Global runs one region’s slow tail moved the provider’s Global p95: Cape Town for Chainstack on Ethereum and BNB Chain, and Sydney and Melbourne for Alchemy on Robinhood Chain and HyperEVM. The reports say that omitting any one of these regions from both providers reverses the full-run win majority.
  • Providers measured. Only Chainstack and Alchemy are compared here. Reports for other providers, including a Chainstack vs QuickNode set, are separate.

Provider results: Chainstack vs Alchemy

1. Chainstack on seven networks

Verdict. Chainstack led Global JSON-RPC on five of six EVM networks and Global newHeads on three of six, and led Ethereum newPendingTransactions in the Global cohort and every region.

Leads. JSON-RPC on Arbitrum, Base, BNB Chain, HyperEVM and Robinhood Chain; newHeads on Base, HyperEVM and Robinhood Chain; Ethereum newPendingTransactions; Sui gRPC in North America and Europe.

Trails. Ethereum newHeads, where Chainstack won 2.3% (101) Global and 0.6% (9) in Asia; Sui gRPC in Asia at 5.4% (1,102); Ethereum JSON-RPC Global at 46.8%.

Regional spread. On Base, Chainstack won JSON-RPC in all four cohorts (62.3% to 67.6%). On HyperEVM, its JSON-RPC share ranges from 90.1% in Europe to 26.0% in North America, a 3.5x spread.

2. Alchemy on seven networks

Verdict. Alchemy led Global JSON-RPC on Ethereum and Global newHeads on Ethereum, Arbitrum and BNB Chain, and led Sui gRPC in the Global cohort.

Leads. Ethereum JSON-RPC and newHeads; Arbitrum and BNB Chain newHeads; Sui gRPC Global and Asia; JSON-RPC in Asia on Ethereum, BNB Chain, HyperEVM and Robinhood Chain.

Trails. Base on every Global measure (33.4% JSON-RPC, 13.2% newHeads); Arbitrum JSON-RPC Global at 23.7%; Ethereum newPendingTransactions at 17.8%.

Regional spread. On Robinhood Chain JSON-RPC, Alchemy’s share ranges from 69.9% in Asia to 27.7% in Europe, a 2.5x spread. On Sui, it ranges from 94.5% in Asia to 9.5% in North America.

Chainstack vs Alchemy JSON-RPC by method

Global cohort. Each cell is the share of the 390 regional p95 comparisons per method that Chainstack won. Above 50% means Chainstack was faster in more locations; the remainder went to Alchemy.

NetworkgetBlockByNumbergetLogs (hash)getLogs (range)getBalancegetTransactionReceiptdebug_traceBlocketh_call
Ethereum49.2%33.6%33.6%60.0%50.5%49.5%51.3%
Arbitrum86.2%83.8%87.4%70.5%82.3%58.5%65.6%
Base66.2%72.6%67.4%70.3%67.7%52.1%70.0%
BNB Chain53.8%45.1%35.1%52.6%57.2%59.7%52.3%
HyperEVM49.7%53.1%53.8%54.4%51.0%54.9%50.0%
Robinhood Chain51.5%54.6%59.2%58.5%53.1%62.8%39.0%

Bold marks Chainstack above 50%. On Ethereum, Alchemy won all 13 runs on Global p95 for eth_getLogs (both variants) and debug_traceBlockByNumber. On BNB Chain, eth_getLogs by range was Chainstack’s weakest method at 35.1%. Arbitrum was above 58% for Chainstack on all seven methods.

Chainstack vs Alchemy RPC by region

JSON-RPC share of regional p95 comparisons won, by cohort (Chainstack / Alchemy):

NetworkNorth America (6)Europe (8)Asia (11)
Ethereum53.5% / 46.5%63.3% / 36.7%24.8% / 75.2%
Arbitrum65.8% / 34.2%81.2% / 18.8%71.1% / 28.9%
Base62.3% / 37.7%62.6% / 37.4%67.6% / 32.4%
BNB Chain58.1% / 41.9%66.5% / 33.5%35.4% / 64.6%
HyperEVM26.0% / 74.0%90.1% / 9.9%38.2% / 61.8%
Robinhood Chain61.4% / 38.6%72.3% / 27.7%30.1% / 69.9%

Chainstack led 13 of the 18 regional JSON-RPC comparisons and Alchemy led 5. All five Alchemy leads are in Asia (four networks) or North America (HyperEVM). Chainstack led in Europe on every network.

WebSocket newHeads, by cohort: Chainstack led 11 of 18 network-region pairs (Base, HyperEVM and Robinhood Chain in all three regions; Arbitrum and BNB Chain in North America only). Alchemy led 7: Ethereum in all three regions, and Arbitrum and BNB Chain in Europe and Asia.

Sui gRPC checkpoints: Chainstack 90.4% in North America and 85.5% in Europe; Alchemy 94.5% in Asia. The reports state no cause for any regional result.

Reliability events recorded

The reports list problems only where they were nonzero.

  • Chainstack: lagging block responses on Ethereum (23 Global, 3 Europe, 2 Asia) and BNB Chain (12 Global, 3 Europe, 2 Asia), and one each on Arbitrum Asia and HyperEVM Asia. In one Base Global batch (Nov 5, run 2), Chainstack’s WebSocket connections failed during warm-up with HTTP 502 errors in Taipei, Osaka and Tokyo; newHeads scoring used the other 27 locations.
  • Alchemy: lagging block responses on HyperEVM (32 Global), Robinhood Chain (3 in Europe, 2 in North America), Base North America (2) and one each on HyperEVM Europe and North America. In one Sui Europe batch (Nov 4, run 1), Alchemy’s Ireland probe recorded zero checkpoints without an error and was excluded.
  • Empty eth_getLogs results appear on both providers (for example Robinhood Chain Global: Chainstack 86 of 1,950, Alchemy 84 of 1,950) and may be valid for blocks with no logs.

Conclusion

On this six-day Chainstack vs Alchemy benchmark, Chainstack won more Global JSON-RPC comparisons on five of six EVM networks, with the widest margin on Arbitrum (76.3% vs 23.7%) and the narrowest on BNB Chain (50.8% vs 49.2%, 46 comparisons). Alchemy led Ethereum JSON-RPC (53.2% vs 46.8%) and Sui gRPC Global (51.4% vs 48.5%).

Read the column that matches the workload:

  • Request-response reads (eth_call, getLogs, getBalance): Chainstack led Global on Arbitrum, Base, HyperEVM and Robinhood Chain; for Ethereum and for any Asia-heavy traffic, check the regional table, where Alchemy led.
  • New-block streaming (newHeads): Chainstack led Base, HyperEVM and Robinhood Chain; Alchemy led Ethereum, Arbitrum and BNB Chain.
  • Mempool streaming on Ethereum (newPendingTransactions): Chainstack led Global and in every region.
  • Sui checkpoints over gRPC: the answer depends on region: Chainstack in North America and Europe, Alchemy in Asia and Global.

Six days of probes from AWS regions is a snapshot, not a trend. Test the endpoints from where your traffic originates before choosing: you can run the same probes yourself with RPC Inspector Pro, or follow live provider scores on Chainstack Compare.

Frequently asked questions

Which provider had lower JSON-RPC latency on each network?

Chainstack won more Global p95 comparisons on Arbitrum, Base, BNB Chain, HyperEVM and Robinhood Chain; Alchemy won on Ethereum. The BNB Chain margin is 1.6 points.

Which provider was faster for WebSocket newHeads?

Chainstack led on Base (85.3%), Robinhood Chain (76.4%) and HyperEVM (71.5%) Global. Alchemy led on Ethereum (97.5%), BNB Chain (73.2%) and Arbitrum (57.8%).

Where did Chainstack lead and trail by region?

Chainstack led JSON-RPC in Europe on all six EVM networks and in North America on five of six (not HyperEVM). It trailed in Asia on Ethereum, BNB Chain, HyperEVM and Robinhood Chain.

How are the results calculated?

For JSON-RPC, lower p95 round-trip time wins each method-region comparison. For WebSocket and gRPC, the provider whose notification arrives first in the same region wins. Ties stay in the denominator only, and no significance test is applied.

Appendix: measurement parameters

ParameterValue
ProvidersChainstack, Alchemy (paid accounts, plan not stated)
NetworksEthereum, Arbitrum, Base, BNB Chain, HyperEVM, Robinhood Chain, Sui
WindowNov 4, 2026 to Nov 9, 2026, 6 days, 13 batches
CohortsGlobal (30 locations), North America (6), Europe (8), Asia (11)
JSON-RPC methodseth_getBlockByNumber, eth_getLogs (hash), eth_getLogs (range), eth_getBalance, eth_getTransactionReceipt, debug_traceBlockByNumber, eth_call
StreamingnewHeads (135 s), newPendingTransactions (35 s, Ethereum), Sui gRPC checkpoints
Request timing15 iterations per 35 s loop, 1,100 ms between starts, 4,000 ms timeout
ProbeOne AWS Lambda per endpoint per location
Ethereum Global sample25,736 RPC requests (Chainstack 12,866; Alchemy 12,870); 902,514 newPendingTransactions arrivals
ScoringLower p95 (JSON-RPC) and first arrival (streaming); no significance test
Author of the reportsCompareNodes.com

Additional resources

SHARE THIS ARTICLE
Customer Stories

DeFiato

Securing a stable environment for platform operations with ease.

Space and Time

Space and Time pioneers Proof of SQL technology at 8x friendlier rates with an Elastic Chainstack data profile for Enterprise.

Kenshi Oracle Network

Contributing towards a swifter Web3 for all with secure, efficient, and robust oracle infrastructure.