GetBlock RPC competitor monitor — 2026-10-13
Criterion of the day: Migration messaging: migrate from Infura/Alchemy/QuickNode, compatibility, zero-downtime migration, backup and failover (migration-backup-failover).
Executive summary
Today’s pass consolidated the full 15-provider universe into a fixed five-dimension migration/resilience schema. The strongest directly captured active-provider story is Chainstack’s combination of migration-guide navigation, endpoint swapping, geo-routing, and Self-Hosted automated failover. dRPC, Tatum, Allnodes, and Lava provide meaningful failover/redundancy evidence but do not expose a complete public zero-downtime migration runbook in the monitored pages.
The most important lifecycle signal remains Blast API’s deprecation notice: it explicitly sends users to Alchemy and promises continuity messaging, but Blast is no longer an active alternative. Grove also states it is winding down; Pocket Network continues with a changed product surface.
Page hashes and broad navigation changes were not treated as market changes. Positive classifications below require criterion-specific first-party text.
Confirmed findings
1. Chainstack exposes the broadest migration/failover public surface in this capture
- The Chainstack documentation exposes “Migrate to Chainstack,” “Migrate with AI,” named provider paths, and a workflow described as swapping endpoints.
- The home page says Global Node routes requests to the nearest available region and describes automatic load balancing across global regions.
- The pricing page describes Self-Hosted deployment with self-healing, automated updates, and failover.
- Limit: no explicit parallel-run, acceptance-test, rollback, or zero-downtime migration procedure was captured. “Uninterrupted” load handling is not equivalent to zero-downtime migration proof.
2. Failover is a clearer competitive claim than migration execution
Direct first-party evidence was captured for:
- dRPC: NodeCore/NodeCloud load balancing, auto-fallbacks, eight geo clusters, and 60+ upstream RPC providers.
- Tatum: Gateway supports adding provider endpoints, load distribution, priority failover, and method routing.
- Allnodes: pricing lists a paid Failover Node add-on, captured at $160/month for Advanced and $320/month for Enterprise; this is volatile and must be verified before sales use.
- Lava Network: home explicitly describes automatic failover across independent providers and chains, with no single point of failure.
- NOWNodes: home/FAQ confirms GEO-balanced routing through an optimal region, but does not explicitly confirm health-triggered fallback in the captured text.
3. Blast-to-Alchemy is an explicit migration event, not a normal feature comparison
The Blast API deprecation notice states that Blast products are no longer available, directs customers to migrate traffic to Alchemy “without interruption,” and offers Alchemy credits with a 15% bonus or a refund. This updates Blast’s matrix row to a discontinued-service transition state. The copy does not include technical cutover steps.
Matrix changes made
- Chainstack: retained migration guidance and failover, strengthened with exact evidence for endpoint swapping, available-region routing, and Self-Hosted automated failover; left zero-downtime cutover unverified.
- dRPC: retained routing/failover and added explicit auto-fallback, 60+ provider, eight-cluster redundancy evidence.
- Tatum: retained failover and added multi-endpoint Gateway redundancy evidence.
- Allnodes: retained failover and added the public failover-node price anchors with a volatility warning.
- Lava Network: retained failover and clarified independent-provider/geographic redundancy.
- NOWNodes: changed from fully unverified to evidenced GEO-balanced routing, without inferring automatic failover.
- Blast API: changed from fully unverified to explicit Blast→Alchemy migration/continuity messaging and marked non-applicable operational cells because the service is deprecated.
- Alchemy: narrowed the prior broad claims to standard JSON-RPC/API compatibility only. A customer “zero downtime” quote was not accepted as zero-downtime migration evidence.
- QuickNode and Infura: removed prior generic compatibility checkmarks because the current evidence did not prove provider-switch compatibility. Generic dedicated infrastructure, high availability, or EVM compatibility is insufficient.
- Moralis: retained migration guidance based on its documentation section, while removing broad compatibility inference not specific to provider migration.
Full-universe evidence and limitations
| Provider | Decision-relevant evidence today | Limitation / extraction gap |
|---|---|---|
| QuickNode | No criterion-specific migration/failover claim verified. | Changelog URL returned 404. Current pricing evidence concerns dedicated infrastructure/SLA, not migration. |
| Alchemy | Official docs expose standard JSON-RPC and chain-specific APIs. | Chains reference and old customer-stories URLs returned 404. Customer zero-downtime quote is not a migration runbook. |
| Infura | Managed API documentation remained accessible. | No named provider migration, cutover, or failover procedure captured. Blog redirected to MetaMask developer news. |
| Chainstack | Migration guide surface, endpoint swapping, geo-routing, self-healing/failover. | No explicit parallel-run/rollback cutover proof captured. |
| Ankr | Docs and enterprise pages accessible. | No criterion-specific public evidence captured. |
| Moralis | Docs expose Migration Guides; home includes API-migration messaging. | No named Infura/Alchemy/QuickNode RPC path or zero-downtime procedure captured. |
| NOWNodes | GEO-balanced routing confirmed. | Legacy Postman docs URL returned 404; health-triggered fallback not explicit. |
| dRPC | Load balancing, auto-fallbacks, geo/provider redundancy. | No public migration runbook captured. |
| Blast API | Deprecated; explicit transition to Alchemy with continuity message. | Not an active provider; no technical cutover instructions in capture. |
| Blockdaemon | Current docs accessible. | RPC and pricing URLs returned 404; no criterion-specific migration/failover evidence captured. |
| Tatum | Gateway failover, load balancing, provider endpoints, method routing. | Supported-blockchains URL returned 404; no migration/rollback runbook captured. |
| Pocket Network / Grove | Grove farewell confirms wind-down; Pocket Network continues. | Grove docs hostname did not resolve; current Pocket product is not directly comparable to the former Grove RPC gateway. |
| LlamaRPC | None verifiable. | Main/chains returned HTTP 526; docs and status hostnames did not resolve. |
| Allnodes | Paid Failover Node option and public price anchors. | Nodes/blog URLs returned 404; activation behavior and SLA mechanics remain unverified. |
| Lava Network | Automatic failover across independent providers/chains. | Developers/network URLs returned 404; no migration procedure captured. |
Cross-provider conclusion
- 9/15 providers have at least one directly evidenced dimension under the normalized schema.
- 0/15 have evidence across all dimensions.
- Six providers have no criterion-specific verified evidence in today’s monitored surface: QuickNode, Infura, Ankr, Blockdaemon, Pocket Network / Grove, and LlamaRPC.
- The largest market-content gap is a public zero-downtime migration playbook covering parallel traffic, method parity, acceptance checks, rollback thresholds, and ownership during cutover.
GetBlock implications
Sales
- Ask prospects whether “backup RPC” means regional balancing, a second backend, or an independent provider; competitors use these concepts interchangeably.
- Use a migration discovery checklist rather than claiming generic drop-in parity: methods, archive/debug, WebSockets, batch behavior, rate limits, transaction broadcast, DNS/endpoint changes, and rollback.
Product marketing
- A differentiated GetBlock page could make the migration contract explicit: source provider, supported chains/methods, parallel-run window, validation checks, rollback criteria, and named support escalation.
- Separate compatibility, zero-downtime cutover, and runtime failover in copy. Current competitor pages often evidence only one of the three.
Content
- Publish provider-specific migration guides (“from Infura,” “from Alchemy,” “from QuickNode”) backed by executable parity tests rather than endpoint-swap claims alone.
- Create a neutral “RPC failover test plan” covering stale heads, partial method failures, WebSocket reconnects, regional loss, rate limiting, and upstream-wide outage.
Source-quality notes
- Direct vendor pages are the basis for positive cells.
- The CompareNodes SOC/ISO reference was reachable but is not relevant to today’s migration/failover criterion; it was not used to classify providers and is not treated as an RPC provider.
- “Changed vs last” flags were treated as collection telemetry only. Most changed pages contained navigation, new-chain, or broad marketing updates rather than proven migration-market changes.
- Claims and prices are volatile; verify before external sales use.
Publishing status supplied by the pre-run
- Encyclopedia pre-publish reported success with 135 URLs and latest daily date 2026-10-13. This report and the cumulative topic brief still require the mandated post-write rebuild and HTTPS QA.
- Agentic routing paper updater reported
ok=true,paper_updated=true, with 61 evidence items, and published the living paper at Agentic routing RPC.