By Eco research. Updated Apr 2026.
​

Blockchain-as-a-Service (BaaS) is a cloud-hosted model in which a provider provisions, monitors, and maintains blockchain nodes on behalf of the customer. Rather than running validator software and peer nodes in-house, teams subscribe to a managed environment and interact with the network through APIs. AWS, Microsoft Azure, Oracle, and IBM all operate BaaS platforms serving enterprise customers across financial services, supply chain, and gaming.
​

Blockchain-as-a-Service is a managed cloud offering that provisions blockchain nodes, handles peer-to-peer networking, and automates protocol upgrades for the subscriber. The customer writes smart contracts and queries the network; the provider handles everything below the application layer including hardware, software patches, monitoring alerts, and key-management infrastructure. Major cloud platforms launched BaaS products between 2018 and 2020.
​

The term maps directly to the broader cloud-services taxonomy: Infrastructure-as-a-Service (IaaS) rents raw compute; Platform-as-a-Service (PaaS) adds a managed runtime; BaaS extends the PaaS model into distributed-ledger networks. AWS Managed Blockchain launched in April 2019, supporting Hyperledger Fabric first and Ethereum later. Microsoft's Azure Blockchain Service followed in 2019, though Microsoft sunset it in September 2021 and redirected customers to partners in the Azure Marketplace. Oracle and IBM retained their dedicated BaaS platforms through 2026.
​

Demand for BaaS grew out of a practical constraint: operating a production blockchain node requires 24/7 uptime, fork-awareness, disk expansion, client upgrades, and operational security practices that most enterprise engineering teams do not maintain in-house. A 2023 Gartner survey found that 58% of enterprise teams cited "operational complexity" as the top barrier to blockchain deployment. BaaS abstracts that complexity behind a service-level agreement.
​

A BaaS platform provisions a blockchain node inside a virtual private cloud, connects it to the target network's peer topology, and exposes HTTP/WebSocket endpoints to the customer. The provider monitors node health, applies protocol upgrades, and scales storage automatically. The customer authenticates via API keys or IAM roles, then deploys contracts and reads chain state through standard JSON-RPC calls.
​

The provisioning flow has three stages. First, the customer selects the network (Ethereum mainnet, Polygon, Hyperledger Fabric, Avalanche) and configuration (full node vs. archive node, redundancy level). Second, the BaaS platform spins up the node, syncs chain history from a snapshot, and routes traffic through load-balanced endpoints. Third, the customer's application points its provider URL at the BaaS endpoint rather than a self-hosted RPC address.
​

Behind the endpoint, the provider runs the full protocol stack: consensus participation (or observation, for non-validator full nodes), mempool management, block propagation, and state storage. Archive nodes retain every historical state root; full nodes prune older states to reduce disk usage. BaaS customers typically choose archive nodes for analytics workloads and full nodes for transaction submission. Hyperledger Fabric introduces an additional component layer (orderer nodes, certificate authorities, chaincode containers) that BaaS platforms like IBM Blockchain Platform manage on the customer's behalf through a Kubernetes-backed control plane.
​

Monitoring is typically delivered through provider-native dashboards. AWS Managed Blockchain pushes node metrics to Amazon CloudWatch; Oracle Blockchain Platform integrates with Oracle Cloud Infrastructure Monitoring. Customers set alerts on block lag, peer count, and response-time thresholds without writing their own monitoring stack.
​

BaaS bundles five core services into a single subscription: node provisioning, protocol upgrades, monitoring and alerting, RPC/WebSocket endpoints, and (on private-network platforms) certificate-authority management. Providers differ on how much of the developer toolchain they include: AWS Managed Blockchain offers only the node layer; IBM Blockchain Platform adds a full console for channel management, chaincode deployment, and identity administration.
​

Specific capabilities vary by provider and network type. Public-network BaaS (Ethereum, Polygon, Avalanche) typically delivers only managed RPC endpoints and node infrastructure. Private-network BaaS (Hyperledger Fabric, Corda) adds governance tooling because those networks require consortium membership management that does not exist on public chains.
​

The provider allocates dedicated virtual machines, installs the consensus client (Geth, Nethermind, or Fabric peer binaries), and manages the startup sequence. On AWS Managed Blockchain, a Fabric network and first member typically provision in under 20 minutes. The provider also handles planned maintenance windows, applying client upgrades without requiring customer intervention.
​

Customers receive HTTPS and WebSocket endpoints compatible with standard Ethereum JSON-RPC (for EVM networks) or Fabric SDK targets (for Hyperledger networks). These endpoints integrate with tooling like Ethers.js, Web3.py, Hardhat, and Truffle without modification. Rate limits and endpoint authentication policies are configurable through the provider's IAM or API-gateway layer.
​

Private blockchain networks require cryptographic identities for every participating organization and peer. IBM Blockchain Platform runs a Fabric CA (certificate authority) per organization and issues enrollment certificates through a web console. This replaces a complex manual PKI setup that can take days to configure correctly. Oracle Blockchain Platform includes a similar managed-CA service alongside its REST proxy, letting developers call chaincode via HTTP without running a Fabric SDK locally.
​

Providers guarantee node uptime through SLAs typically ranging from 99.9% to 99.99% depending on redundancy configuration. Monitoring integrates into cloud-native alerting stacks (CloudWatch, Azure Monitor, OCI Monitoring). Out-of-the-box metrics include block height lag, transaction throughput, peer count, and storage utilization. This eliminates the need for a dedicated DevOps function to maintain node observability.
​

Four enterprise providers dominate the managed blockchain infrastructure market: AWS, Oracle, IBM, and the Azure Marketplace ecosystem (following Microsoft's 2021 direct-service sunset). They differ on supported networks, deployment model, pricing structure, and console depth. The table below summarizes the major dimensions for teams evaluating options as of Q2 2026.
​

AWS is the most common entry point for teams already on Amazon's cloud because it requires no additional vendor relationship. IBM Blockchain Platform targets enterprises that need deep Fabric governance tooling and optionally want to deploy the same platform on-premises through IBM Cloud Pak for Business Automation. Oracle's platform is the most common choice for organizations running Oracle E-Business Suite or Oracle ERP, since it ships with pre-built connectors to Oracle Autonomous Database.
​

Microsoft retired Azure Blockchain Service in September 2021 after concluding that partner-led BaaS (through the Azure Marketplace) served enterprise customers better than a first-party product. Microsoft's official retirement FAQ explicitly named Kaleido and ConsenSys Quorum as recommended migration paths.
​

BaaS, self-hosted nodes, and RPC-as-a-Service (RaaS) each sit at a different point on the control-vs.-convenience spectrum. BaaS delivers SLA-backed managed infrastructure but introduces vendor lock-in. Self-hosting maximizes control at the cost of ongoing node operations. RaaS providers like Alchemy and Infura serve public-chain workloads at lower per-request cost than BaaS for most throughput levels.
​

The meaningful distinctions come down to three variables: network type, throughput, and compliance posture. For private Hyperledger Fabric networks, BaaS from IBM or Oracle is nearly always the right choice because self-hosting Fabric requires Kubernetes expertise that most enterprise teams lack. For public Ethereum or Polygon workloads under 50 million requests per month, RaaS from Alchemy or Infura costs less and delivers equivalent performance. For workloads that require data residency within a specific cloud region and tight IAM integration, BaaS inside a team's existing cloud account is the clearest fit.
​

"Enterprise blockchain projects fail primarily at the operations layer, not the smart-contract layer. Having a managed platform that handles node upgrades and monitoring removes the single biggest source of production incidents." — Steve Cerveny, Co-founder at Kaleido

Teams evaluating public-network BaaS against RaaS should compare the Alchemy node infrastructure overview against AWS Managed Blockchain's Ethereum offering on two dimensions: archive data access (BaaS archive nodes retain full historical state; some RaaS providers charge separately) and VPC integration (BaaS nodes live inside the customer's cloud account, which simplifies private connectivity to on-premises systems).
​

Self-hosting remains the choice for teams with hard requirements on client diversity (running a minority client like Besu or Nimbus to reduce network centralization risk) or teams that need to participate in validator sets rather than just observe the network. BaaS platforms do not offer validator staking participation for Ethereum's proof-of-stake consensus; they provision full nodes only. For rollup deployment, teams that want to run their own sequencer or full L2 node require self-hosting or a specialized rollup-infrastructure provider, not a general BaaS platform.
​

BaaS sees active deployment across four broad domains: enterprise supply chain tracking, financial-services settlement, gaming item ownership, and DeFi infrastructure. In each case the operational benefit is the same: engineering teams interact with smart-contract or ledger APIs and delegate node upkeep to the provider. The use cases diverge primarily in which network type and which provider they favor.
​

Maersk and IBM jointly built TradeLens, a Hyperledger Fabric network for container logistics, on IBM Blockchain Platform. The network at peak tracked over 60% of global containerized trade, connecting 175 organizations across ports, shipping lines, and customs authorities. (TradeLens was sunset in December 2022 after Maersk's industry-wide adoption strategy did not materialize, but the technical architecture remained a reference case for enterprise BaaS deployments.) The key lesson: BaaS enabled fast onboarding of consortium members without each organization running its own Fabric infrastructure.
​

Banks running tokenized bond or repo settlement pilots frequently use AWS Managed Blockchain or Oracle Blockchain Platform because both integrate with existing cloud-native compliance tooling. The Bank for International Settlements' Project Mariana (2023) demonstrated that shared tokenized FX infrastructure could settle in 90 seconds vs. the two-day standard settlement cycle. While Project Mariana used a custom DeFi stack, its findings accelerated enterprise interest in managed blockchain infrastructure that could support similar settlement architectures.
​

Game studios use BaaS to anchor NFT item records without building internal blockchain operations. AWS Managed Blockchain supports Amazon Managed Blockchain Access for Ethereum, letting game backends verify token ownership through standard RPC calls routed inside a VPC. Studios avoid the operational burden of running an Ethereum full node while retaining the ownership-verification guarantees of the public ledger.
​

Development teams building DeFi protocols often use BaaS-adjacent RaaS providers (Alchemy, Infura, QuickNode) for staging and production RPC access, but some organizations with strict data-residency requirements or high-compliance audit needs prefer BaaS nodes inside their own cloud account. Alchemy's 2024 developer survey reported that 23% of enterprise-classified DeFi teams used a managed node service inside their primary cloud account vs. a shared RaaS endpoint, citing audit trail requirements as the primary driver.
​

BaaS introduces vendor dependencies and architectural constraints absent in self-hosted deployments. Provider lock-in, narrow network coverage, pricing that exceeds self-hosting at scale, and the absence of validator participation are the four most common production limitations. Teams should weigh each against their compliance posture and operational capacity before committing to a BaaS provider.
​

Provider lock-in is the most cited concern. BaaS endpoints use proprietary authentication and monitoring integrations; migrating a production application from AWS Managed Blockchain to Oracle or a self-hosted node requires re-configuring IAM, updating endpoint URLs, migrating CloudWatch dashboards, and often rewriting infrastructure-as-code templates. AWS and Oracle use sufficiently different control planes that the migration is a meaningful engineering project. No BaaS platform exposes raw virtual-machine access or lets the customer swap consensus clients; Geth runs on AWS Managed Blockchain because AWS chose it, not because the customer did.
​

Network coverage is narrow. AWS Managed Blockchain supports two networks: Ethereum and Hyperledger Fabric. A team building on Solana, Avalanche, Arbitrum, or Base cannot use AWS BaaS. Oracle supports only Hyperledger Fabric. IBM supports Fabric with optional on-premises deployment. Teams on non-covered networks default to RaaS providers or self-hosting. For cross-chain applications that move assets between networks via blockchain bridges, BaaS covers only the chain-connectivity layer and provides no bridging orchestration. A team building a multi-chain stablecoin payment system spanning Ethereum, Arbitrum, and Polygon cannot consolidate all three chains onto a single BaaS contract; they end up mixing AWS Managed Blockchain for Ethereum with RaaS for the L2 chains.
​

Pricing can exceed self-hosting at scale. AWS Managed Blockchain charges per node-hour, per peer-member, and per I/O request; a single-member two-peer Fabric network runs roughly $800 to $1,200 per month before data transfer. A team running equivalent infrastructure on a self-managed EC2 cluster could run the same workload for $300 to $500 per month but with full operational responsibility. At high throughput, RaaS providers typically undercut both options for public-network workloads: Alchemy's Growth plan handles 300 million compute units per month for $199, which covers most DeFi protocol workloads far below the equivalent BaaS node cost.
​

BaaS platforms do not offer Ethereum proof-of-stake validator participation. Customers provision full nodes that sync the chain and serve RPC requests but do not propose or attest blocks. Teams that want staking yield or validator-level network participation must use a staking service or self-hosted validator setup. This is a meaningful limitation for teams studying Ethereum's node-and-client documentation and considering whether managed infrastructure can substitute for a full validator deployment: it cannot.
​

Data portability is a related risk. Archive data on BaaS nodes belongs to the provider's storage layer. If a team cancels its contract, re-syncing a new archive node from genesis on Ethereum takes days and requires significant disk provisioning. Teams running time-sensitive analytics or forensic queries should maintain a secondary archive node or export historical data independently before any provider transition.
​

AWS Managed Blockchain supports two networks as of Q2 2026: Ethereum (public mainnet and testnets) and Hyperledger Fabric (private consortium networks). Solana, Avalanche, Arbitrum, Polygon, and other EVM-compatible L2 chains are not available on AWS BaaS. Teams needing those networks typically use RaaS providers such as Alchemy, QuickNode, or Infura instead.
​

They are related but distinct categories. BaaS platforms (AWS, IBM, Oracle) provision dedicated nodes inside the customer's cloud account with SLA-backed uptime and deep IAM integration. RaaS providers like Alchemy and Infura run shared or dedicated node pools and deliver access via API key, typically with faster setup, lower cost at moderate volumes, and broader chain coverage but without in-account data residency or enterprise compliance integrations.
​

Yes. Private permissioned networks are the primary use case for IBM Blockchain Platform and Oracle Blockchain Platform, both of which are built on Hyperledger Fabric. These platforms manage consortium membership, certificate authorities, channel governance, and chaincode lifecycle on behalf of the customer. Public-chain BaaS (AWS Ethereum) handles permissionless public networks only and does not provide consortium governance tooling.
​

Microsoft's 2021 Azure Blockchain Service sunset is the reference case. Microsoft gave customers 12 months' notice and explicitly recommended migrating to Kaleido, ConsenSys Quorum, or other Azure Marketplace partners. Teams that had hard-coded endpoint URLs and proprietary monitoring integrations faced a full re-architecture. The lesson most enterprises drew was to abstract BaaS endpoints behind a configuration layer so that endpoint URLs can be swapped without application code changes. See Microsoft's official retirement FAQ for the migration guidance they published.
​

BaaS is a workable choice for game studios that need Ethereum full-node access to verify NFT ownership and token state inside a VPC. AWS Managed Blockchain's Ethereum offering supports the standard JSON-RPC interface used by most NFT verification libraries. The main limitation is throughput: high-concurrency games with millions of daily active users may need dedicated or load-balanced RaaS endpoints rather than a single BaaS node to handle peak verification request volume without latency impact.
​

Eco's stablecoin routing layer spans 15 chains and relies on onchain RPC access to observe mempool state, estimate gas, and confirm transaction finality. Enterprise teams integrating the Eco API connect their application to Eco through standard HTTPS; Eco manages the underlying node infrastructure. The BaaS-vs.-self-hosting trade-offs described above still apply when teams build their own adjacent application layer.
​

For teams building treasury automation or payment flows, the node infrastructure decision is largely separate from the routing-and-settlement decision. BaaS or RaaS handles the raw RPC layer; Eco handles the routing logic, bridge selection, and settlement finality across chains. The two layers are complementary rather than substitutes. A team running its own analytics node on AWS Managed Blockchain and using Eco's API for settlement execution is a common pattern in enterprise stablecoin payment deployments.
​

Sources and methodology. BaaS platform feature sets verified against AWS Managed Blockchain documentation, Oracle Blockchain Platform documentation, and IBM Blockchain Platform documentation as of April 2026. Microsoft Azure Blockchain Service retirement confirmed via Microsoft's official retirement FAQ. BaaS adoption barriers from Gartner 2023 enterprise blockchain survey. Figures and product status reflect Q2 2026 provider offerings.
​

Teams building cross-chain payment or settlement applications can learn how Eco's routing layer handles multi-chain stablecoin transfers without requiring each team to operate their own node infrastructure at eco.com.
​