SQD vs Chainstack

SQD vs Chainstack: RPC nodes or indexed data?

225+ networks · Sub-second real-time · Zero egress, no per-call metering

Chainstack is a managed node provider: RPC endpoints across 70+ protocols billed in request units, with dedicated and self-hosted node options. SQD is an indexed-data layer that streams decoded, validated data across 225+ networks from one Portal endpoint without per-call metering. Chainstack is the better fit for live RPC and transaction submission; SQD is the better fit for high-volume historical backfills and real-time streaming. Note that Chainstack's Subgraphs indexing product is being sunset in early 2026.

Pick SQD if

  • You need decoded, validated data across many networks from one streaming endpoint rather than raw JSON-RPC you decode yourself.
  • You run large historical backfills or high-volume reads where per-call RU metering against archive nodes gets expensive.
  • You want an open-source, self-hostable stack and a free public Portal with no API key, no card, and no token.

Pick Chainstack if

  • You need to submit transactions and have direct JSON-RPC node access; SQD is read-side only and cannot write to chains.
  • You want low-latency trading infrastructure such as Solana Yellowstone gRPC streaming or a Trader Node mempool feed.
  • You want dedicated or isolated node compute in specific cloud regions (AWS, Azure, GCP, Virtuozzo across USA, UK, EU, Asia) with an enterprise SLA and SOC 2 Type II.

Two different layers

Where SQD and Chainstack sit in your data stack

Chainstack sits at the node-provider layer: it runs RPC nodes (shared, dedicated, global, and a self-hosted control plane) that return raw JSON-RPC responses you query block by block. SQD sits one layer up at the indexed-data layer: a Portal API that streams decoded, validated historical and real-time data across 225+ networks without per-call RPC metering. The two are complementary; many teams keep an RPC provider like Chainstack for live transaction submission and node access while using SQD to backfill and stream large volumes of historical data.

  1. Apps & products
    Wallets Tax Payments KYC RWA
  2. Intelligence
    Comparison coming soon Comparison coming soon Comparison coming soon Comparison coming soon Comparison coming soon
  3. Protocol analytics
    Dune Comparison coming soon
  4. Indexed data
  5. Our focus
    Read-side infrastructure
    SQD decentralized, validated, multi-chain at source
  6. Node providers

SQD optimizes for

  • Decoded data at source: one endpoint, 225+ chains, queryable directly
  • Multi-chain in one endpoint, not per-chain RPC metered by credits or request units
  • A TypeScript SDK for custom indexing, plus a streaming Portal API
  • Validated data with 6-step verification at ingestion
  • Open source (Portal node AGPL-3.0) and self-hostable

Chainstack optimizes for

  • Managed JSON-RPC nodes across 70+ protocols, billed in request units
  • Transaction submission and direct node access (shared, dedicated, self-hosted)
  • Low-latency trading: Solana Yellowstone gRPC and a Trader Node mempool feed
  • Dedicated node compute pinned to specific cloud regions with a 99.99% SLA

At a glance

Side-by-side comparison

Dimension SQD Chainstack
Coverage & Performance
Networks / protocols Chainstack count from its protocols page; SQD count from /chains/ catalogue. 225+ networks across EVM, Solana, Substrate, Bitcoin, Hyperliquid 70+ protocols (EVM, Solana, Sui, Aptos, Bitcoin, Hyperliquid and more)
Data model Decoded, indexed historical + real-time data via one Portal endpoint Raw JSON-RPC responses per node/chain; no unified cross-chain query API
Real-time latency Different shapes: SQD streams decoded data; Chainstack serves RPC. Sub-second hot path; independent benchmark 27ms P50 / 48.5ms P90 Real-time RPC; Solana Yellowstone gRPC streaming and low-latency Trader Node mempool available
Historical backfill Archive request = 2 RU per Chainstack pricing. Stream entire history from one endpoint, zero egress Archive node access; billed at 2 RU per request, so large backfills consume RU quickly
Transaction submission / writes Chainstack RPC handles eth_sendRawTransaction and Warp Transactions; SQD is read-side only.
Architecture & openness
Layer in the stack Indexed-data layer Node-provider layer (RPC)
Decoded data out of the box Chainstack returns raw RPC; decoding is the developer's responsibility.
Cross-chain query interface One Portal API and one contract across SQD; Chainstack endpoints are per-chain.
Open source Chainstack core/control plane proprietary; open-source SDKs/tooling only. Pipes SDK MIT, Portal node AGPL-3.0
Self-hostable Chainstack Self-Hosted is a Kubernetes-native control plane; SQD self-host is fully open source. Yes (Squid SDK + Portal node) plus managed cloud Self-Hosted control plane on your own infra (proprietary software)
Decentralized network SQD: worker operators paid in SQD. Chainstack: centralized managed service.
Data validation 6-step cryptographic validation at ingestion Not published (relies on underlying node clients)
Economics
Pricing model Chainstack RU model from pricing page. Predictable, network-based; not per-call metered Request-unit metering (1 RU full node, 2 RU archive) plus flat-fee node options
Free tier Chainstack free tier capped by RU; SQD public Portal uncapped at the free tier. Free public Portal: no API key, no card Developer plan: $0, 3M RU/month, 25 RPS, 1 project
Egress fees Zero egress No separate egress; consumption governed by RU and RPS limits
Token required for production Neither requires holding a token to use in production. Chainstack accepts crypto payment but does not require its own token.
Products
Dedicated indexing / subgraphs Status from Chainstack Subgraphs sunset notice. Squid SDK + Pipes SDK for custom indexers Subgraphs product being sunset (no new subgraphs after Jan 5, 2026; disabled Feb 1, 2026; migrating to partner Ormi)

Facts verified June 2026 against Chainstack’s public docs and product pages. Every cell value is cited in the Sources section at the bottom of this page.

Decision framework

Choose based on your workload

When SQD is the better choice

  • High-volume historical backfills where Chainstack's 2-RU-per-archive-request metering adds up fast.
  • Multi-chain projects that want one decoded query interface across 225+ networks instead of per-chain RPC endpoints.
  • Teams that need cryptographically validated data and an open-source, self-hostable pipeline.
  • Cost-sensitive teams that benefit from a free public Portal and network-based (not per-call) pricing.

When Chainstack is the better choice

  • Live transaction submission and write paths (eth_sendRawTransaction, Warp Transactions).
  • Low-latency trading: Solana Yellowstone gRPC streaming and Trader Node mempool access.
  • Dedicated/isolated node compute pinned to specific cloud regions with a 99.99% SLA.
  • Teams that explicitly want managed RPC node access rather than an indexed-data layer.

Most production teams use both: Chainstack for its layer, and SQD for the decoded data layer above it.

What changes, what stays

Where SQD replaces Chainstack, and where you keep both

SQD does not replace Chainstack wholesale; it replaces the read-heavy half of the workload. Point your historical backfills and large data reads at SQD's Portal, which streams decoded, validated data across 225+ networks without per-call metering or egress fees. Keep Chainstack for what a node provider does best: live JSON-RPC, transaction submission, and any low-latency trading paths such as Yellowstone gRPC or the Trader Node mempool. Note separately that Chainstack's Subgraphs indexing product is being sunset in early 2026, so any custom indexing on it needs a new home regardless; SQD's Squid SDK and Pipes SDK are one option there.

SQD replaces

  • Archive-RPC reads and large historical backfills (metered at 2 RU per request)
  • Per-chain data fetching and manual decoding of raw RPC responses
  • Custom indexing previously built on the sunsetting Chainstack Subgraphs product

You keep Chainstack for

  • Live JSON-RPC endpoints and transaction submission
  • Solana Yellowstone gRPC streaming and Trader Node low-latency mempool
  • Dedicated node compute in specific cloud regions with enterprise SLA

The matrix above breaks the comparison down dimension by dimension. Every claim is cited in Sources.

The full SQD toolkit

What you get with SQD

SQD offers four products that share the same underlying data lake. Pick the layer that fits your workload.

Product What it does Best for
Portal API Sub-second multi-chain query API Real-time apps, trading, agents
Squid SDK TypeScript indexer framework Custom data products, GraphQL APIs
Pipes SDK Streaming ETL pipelines Warehouses, analytics, batch loads
SQD Network Decentralized data lake Open access, no vendor lock

Need a custom API? Build it with SQD. When a fixed, hosted endpoint doesn't fit, the Squid SDK and Pipes SDK let you build your own indexer or streaming pipeline in TypeScript over any of 225+ chains, shaped to your schema. The Squid SDK deploys to SQD Cloud, or self-host either.

225+

Networks indexed

27ms

Median Portal response

6-step

Validation at ingestion

Frequently asked questions

Is SQD an alternative to Chainstack?
For historical backfills and real-time data streaming, yes. SQD's Portal API streams decoded, validated data across 225+ networks from one endpoint without per-call metering, which replaces the heavy archive-RPC reads many teams run against Chainstack. For live transaction submission and node access, Chainstack remains complementary: SQD is read-side only and does not submit transactions.
What is the difference between Chainstack and SQD?
Chainstack is a node provider: it runs RPC nodes across 70+ protocols and returns raw JSON-RPC responses, billed in request units. SQD is an indexed-data layer: it serves decoded historical and real-time data across 225+ networks through one streaming Portal API. Chainstack gives you a node to talk to; SQD gives you queryable, decoded data.
Is Chainstack cheaper than SQD for indexing?
It depends on workload. Chainstack bills in request units (1 RU per full-node request, 2 RU per archive request), so large historical backfills consume RU quickly because every block read is metered. SQD offers a free public Portal with no API key or card, and paid tiers that are network-based rather than per-call metered, which tends to favor high-volume historical and streaming workloads. See SQD pricing and the Chainstack pricing page.
Does Chainstack still offer Subgraphs?
Chainstack's Subgraphs indexing product is being sunset. Per Chainstack's notice, no new subgraphs can be created after January 5, 2026, and the service is disabled on February 1, 2026, with existing subgraphs migrating to a partner indexer (Ormi). For ongoing custom indexing, SQD provides the open-source Squid SDK and Pipes SDK.
Can I self-host Chainstack or SQD?
Both offer self-hosting, but differently. Chainstack Self-Hosted is a proprietary Kubernetes-native control plane that deploys and operates nodes on infrastructure you control. SQD is self-hostable end to end with open-source components (Squid SDK and the Portal node under AGPL-3.0), alongside managed SQD Cloud and a free public Portal.
Is Chainstack or SQD better for Solana data?
Chainstack offers Solana RPC plus Yellowstone gRPC streaming and a low-latency Trader Node mempool, which suit trading and transaction submission. SQD covers Solana in its 225+ networks with decoded, streamed instruction and transaction data and a sub-second hot path, which suits analytics, dashboards, and large historical backfills. Many teams use both.
Does SQD or Chainstack require an API key?
Chainstack requires an account and issues per-endpoint credentials; usage is metered in request units. SQD's free public Portal requires no API key and no card, and no token is required for production use. Paid SQD tiers add predictable, network-based pricing.
Sources & methodology (9 citations, verified June 2026)

Every comparison cell and statement about Chainstack is anchored to Chainstack’s own public documentation. Spotted something stale? Let us know.

Coverage & protocols

  • Chainstack supports 70+ blockchain protocols including Ethereum, Solana, Sui, Aptos, Bitcoin, Hyperliquid, plus newer additions MegaETH, Tempo, Monad [Chainstack protocols page]
  • Homepage lists Bitcoin among supported chains and states a 99.99% SLA and SOC 2 Type II [Chainstack homepage]

Pricing & request units

  • Developer $0 (3M RU, 25 RPS) / Growth $49 (20M RU, 250 RPS) / Pro $199 (80M RU, 400 RPS) / Business $499 (200M RU, 600 RPS) / Enterprise from $990 (400M RU, unlimited RPS); 1 RU per full-node request, 2 RU per archive request; overage $20 down to $5 per 1M RU; crypto payment on all plans [Chainstack pricing page]
  • Dedicated node compute from $0.50/hour, storage from $0.01 per 20 GB/hour; AWS, Azure, GCP, Virtuozzo across USA, UK, Germany, Netherlands, Singapore, Japan [Chainstack Dedicated Node page]

Products & architecture

  • Self-Hosted is a Kubernetes-native control plane for deploying, operating, and monitoring nodes on infrastructure you control (own cloud, on-premise, or partner hardware) [Chainstack Self-Hosted page]
  • Node types: Global Node, Unlimited Node, Dedicated Node, Trader Node, Self-Hosted; Yellowstone gRPC and Warp Transactions; 99.99% SLA and SOC 2 Type II [Chainstack homepage]

Subgraphs sunset

  • Subgraphs being sunset: no new subgraphs after Jan 5, 2026; service disabled Feb 1, 2026; existing subgraphs migrate to partner Ormi (one complimentary month Feb 1 to Mar 1, 2026) [Chainstack Subgraphs docs]

SQD facts

  • SQD: 225+ networks, one Portal endpoint, streaming HTTP, zero egress, sub-second hot path, 27ms P50 / 48.5ms P90; free public Portal, no token required [SQD Portal]
  • SQD open source: Pipes SDK MIT, Portal node AGPL-3.0 [SQD GitHub]

Decoded data on 225+ chains, one endpoint

Free public Portal, open-source SDK, and 225+ chains indexed from day one.

Indexing more than 100M events a day or need a dedicated portal? Talk to sales.