The Graph's Hosted Service excels at developer velocity and operational simplicity because it is a fully managed, centralized service run by Edge & Node. For example, it offers a 99.9%+ uptime SLA, handles all node operations, and provides instant, free querying for development—critical for rapid prototyping of dApps like Uniswap or Aave in their early stages. This eliminates the overhead of managing indexer infrastructure, allowing teams to focus solely on subgraph development.
LABS
Comparisons
The Graph's Hosted Service vs Decentralized Network: The Definitive Migration Path
A technical analysis for CTOs and protocol architects comparing The Graph's free hosted service against its decentralized mainnet. We evaluate performance, cost, security, and provide a clear framework for migration timing.
Chainscore © 2026
introduction
THE ANALYSIS
Introduction: The Centralization-Decentralization Pivot
A data-driven comparison of The Graph's Hosted Service and its Decentralized Network, framing the migration as a strategic infrastructure choice.
The Graph's Decentralized Network takes a different approach by distributing the indexing and querying workload across a permissionless network of independent Indexers, Curators, and Delegators. This results in a trade-off: while introducing complexity and variable query costs (e.g., 0-100+ GRT per 1k queries), it delivers censorship resistance, data provenance via cryptographic proofs, and alignment with Web3's core values. The network secures over $2.5B in Total Value Locked (TVL) in its staking contracts, underpinning its economic security.
The key trade-off: If your priority is time-to-market, predictable costs (zero during dev), and a hands-off operational model, the Hosted Service is the pragmatic choice. If you prioritize decentralized security, protocol resilience, and building on credibly neutral infrastructure for a production mainnet dApp, the Decentralized Network is the strategic, long-term bet. The migration path is well-documented, allowing projects to start on the Hosted Service and transition as their decentralization requirements mature.
tldr-summary
The Graph Hosted Service vs Decentralized Network
TL;DR: Key Differentiators at a Glance
A quick-scan breakdown of core trade-offs to inform your migration strategy.
02
Hosted Service: Cost Predictability
Zero direct query costs. Queries are served from a shared, subsidized cache. This matters for bootstrapped projects or applications with unpredictable, high-volume query patterns where variable GRT costs are a concern.
03
Decentralized Network: Censorship Resistance
Decentralized Indexers & Curators: Data is served by a permissionless network of over 700 Indexers. This matters for mission-critical DeFi protocols (like Uniswap, Balancer) requiring uptime guarantees and resistance to centralized takedowns.
700+
Active Indexers
04
Decentralized Network: Data Integrity & Incentives
Cryptoeconomic Security: Indexers stake GRT and are slashed for malfeasance; Curators signal on quality subgraphs. This matters for high-value financial data where tamper-proof indexing is non-negotiable.
05
Decentralized Network: Long-Term Viability
Aligned with Web3 ethos. The hosted service is being sunset (Q1 2025). Migration is inevitable for production apps. This matters for any project planning beyond 2024 that requires a sustainable data layer.
06
Decentralized Network: Query Flexibility
Performance SLAs & Customization: You can choose Indexers based on performance, price, and geography. This matters for enterprise applications needing guaranteed latency (< 500ms p95) or specialized data pipelines.
HEAD-TO-HEAD COMPARISON
The Graph: Hosted Service vs Decentralized Network
Key operational and economic metrics for infrastructure decision-making.
| Metric | Hosted Service | Decentralized Network |
|---|---|---|
Cost Model | Free Tier + Usage-Based | GRT Query Fees + Indexer Stakes |
Uptime SLA | Decentralized (No Single Point) | |
Data Freshness | ~1 block | ~1 block |
Protocol Governance | The Graph Foundation | GRT Token Holders & Delegators |
Query Throughput Limit | ~100 req/sec (Free Tier) | Unlimited (Market-Based) |
Migration Effort | Subgraph Deployment Only | Subgraph Curation & Indexer Selection |
Primary Use Case | Prototyping & Early Growth | Production & Censorship Resistance |
pros-cons-a
PROS AND CONS
The Graph Hosted Service vs Decentralized Network: Migration Path
Key strengths and trade-offs for CTOs evaluating a migration from the hosted service to The Graph's decentralized mainnet.
01
Hosted Service: Operational Simplicity
Zero infrastructure overhead: Managed by The Graph Foundation, requiring no node operation or GRT token management. This matters for teams with limited DevOps resources or those in rapid prototyping phases, allowing them to query data via a simple API key.
02
Hosted Service: Cost Predictability
Fixed, fiat-based billing: No exposure to GRT price volatility. This matters for startups with strict, predictable OpEx budgets, as costs scale linearly with query volume without the complexity of crypto payments.
03
Decentralized Network: Censorship Resistance
Unstoppable data availability: Indexers and curators are globally distributed, ensuring subgraphs remain queryable without a central point of failure. This matters for DeFi protocols like Uniswap or Aave that require guaranteed uptime and neutrality.
04
Decentralized Network: Protocol Alignment & Incentives
Direct participation in Web3 economics: Earn query fees and indexing rewards in GRT. This matters for protocols building their own subgraphs, as they can stake GRT to curate their data and attract higher-quality indexers, improving performance.
05
Hosted Service: Strategic Risk
Scheduled sunset: The hosted service is a deprecated product, with a defined migration timeline. This matters for long-term projects, as building on it creates technical debt and requires a future, potentially complex, migration effort.
06
Decentralized Network: Operational Complexity
Token economics & management: Requires handling GRT for payments, staking, and delegation. This matters for traditional enterprises or teams unfamiliar with crypto wallets, gas fees, and managing treasury exposure to a volatile asset.
pros-cons-b
Hosted Service vs Decentralized Network
The Graph Decentralized Network: Pros and Cons
Key strengths and trade-offs for CTOs evaluating the migration from the managed Hosted Service to the permissionless Decentralized Network.
01
Hosted Service: Pros
Managed Infrastructure: Zero operational overhead for Indexer nodes, curation, and query gateways. This matters for teams that need to ship fast without managing GRT staking or node operations.
Predictable Costs: Simple, usage-based billing in USD (via credit card). No exposure to GRT token price volatility, which is critical for stable project budgeting.
Proven Reliability: Serves over 1 trillion queries monthly for protocols like Uniswap, Aave, and Balancer. Ideal for production applications requiring a proven, battle-tested service level.
02
Hosted Service: Cons
Centralized Dependency: Relies on a single entity (The Graph Foundation) for uptime and policy. A single point of failure that contradicts decentralization principles for protocols like Lido or Compound.
Limited Customization: Fixed query performance and indexing rules. Not suitable for projects needing bespoke indexing logic or subgraph-to-subgraph calls.
Sunset Risk: The service is a transitional product. Long-term roadmap mandates migration to the Decentralized Network, creating eventual technical debt.
03
Decentralized Network: Pros
Censorship Resistance: Queries are served by a permissionless network of Indexers (e.g., Figment, Pinax). Essential for DeFi and DAO data that must remain accessible under any circumstances.
Economic Alignment: Uses GRT for payments and staking, aligning incentives between Indexers, Curators, and Delegators. This creates a sustainable ecosystem for long-term data integrity.
Performance & Customization: Can leverage Indexer-specific features like faster indexing, multi-chain support, and custom firehose setups. Critical for high-frequency dApps like perpetuals exchanges or NFT marketplaces.
04
Decentralized Network: Cons
Operational Complexity: Requires managing GRT for query payments, staking, and delegation. Adds significant DevOps and treasury management overhead.
Cost Volatility: Query costs fluctuate with GRT price and network demand. Budget forecasting becomes challenging compared to fixed USD pricing.
Maturity & Support: While robust, the network has fewer integrated tools (e.g., dashboarding, alerting) than the Hosted Service. Teams may need to build or adapt their own monitoring, increasing initial setup time.
CHOOSE YOUR PRIORITY
When to Choose: Decision Framework by Team Profile
The Graph Hosted Service for Speed & Simplicity
Verdict: The clear choice for rapid prototyping and production apps where developer velocity is paramount. Strengths: Zero infrastructure overhead, instant subgraph deployment, and predictable, low-latency query performance. The managed service handles indexing node orchestration, database scaling, and query gateway load balancing. This is ideal for hackathons, MVPs, or teams without dedicated DevOps resources. Projects like Uniswap V3 and Aave initially leveraged the hosted service for its reliability.
The Graph Network for Speed & Simplicity
Verdict: Not ideal. The decentralized network introduces complexity that slows initial development. Trade-offs: Requires managing GRT curation, selecting and monitoring Indexers, and dealing with potential indexing disputes. Query performance can vary based on Indexer selection and network congestion. The migration path from hosted service to network is well-documented but adds a step.
THE GRAPH HOSTED SERVICE VS DECENTRALIZED NETWORK
Migration Path: Step-by-Step Guide to Decentralization
A technical guide for engineering leaders evaluating the migration from The Graph's centralized Hosted Service to its decentralized Mainnet. This comparison covers cost, performance, security, and the practical steps for a successful transition.
The Decentralized Network introduces a predictable, usage-based billing model, while the Hosted Service is free but unsustainable. Migrating to the network requires budgeting for GRT tokens to pay Indexers for queries and indexing. Costs are determined by a decentralized marketplace, with typical query fees being fractions of a cent. The Hosted Service, while currently free, is being phased out and offers no service-level agreements (SLAs) or long-term reliability, representing an operational risk for production applications.
verdict
THE ANALYSIS
Final Verdict and Strategic Recommendation
A data-driven breakdown to guide your infrastructure migration from The Graph's Hosted Service to its Decentralized Network.
The Graph's Hosted Service excels at providing a stable, zero-cost entry point for rapid prototyping and early-stage dApps. Its fully managed infrastructure, with a 99.9%+ uptime SLA, eliminates operational overhead, allowing teams like Uniswap V3 and Aave to scale their initial subgraphs without managing indexers or staking GRT. The primary constraint is its sunset timeline, with query service ending in Q3 2024, forcing a strategic decision.
The Decentralized Network takes a different approach by prioritizing censorship resistance, data integrity, and long-term protocol alignment. This results in a trade-off of increased initial complexity—requiring GRT for curation and query fees—for superior decentralization and economic security. The network now indexes over 40 blockchains and secures ~$2B in stake (TVL), making it the default for production-grade applications like Livepeer and Decentraland that require unstoppable data feeds.
The key trade-off: If your priority is minimizing time-to-market and development cost for a new project, use the Hosted Service as a temporary launchpad. If you prioritize production resilience, verifiable data, and alignment with web3 principles, migrate to the Decentralized Network immediately. For teams with existing subgraphs, the recommended path is to use the Graph CLI's built-in migration tools to deploy to the decentralized network, leveraging the same subgraph manifest for a streamlined transition.
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall