What happens to an indexing protocol that already processes data for most of Web3 when the blockchain infrastructure market is projected to grow from $33 billion to $393 billion by 2030, the agentic AI economy is forecast to reach $47 billion in the same timeframe, and DTCC has already demonstrated real-time collateral settlement on blockchain rails using technology from its ecosystem?
It either evolves or gets replaced.
The Graph's 2026 technical roadmap, published this week, is the protocol's answer to that question. The roadmap outlines a structural pivot from a single-product indexing service into a modular, multi-service data platform, with six specialized products targeting blockchain developers, AI agents, DeFi infrastructure, and regulated financial institutions. It is built on Horizon, the protocol upgrade that went live in December 2025, which decoupled The Graph's core infrastructure from subgraphs and opened the architecture to support any data service under one shared staking, payment, and security framework.
The numbers behind The Graph's existing reach are substantial. The network has processed over 1.27 trillion queries since launch. Quarterly query volume hit a record 6.49 billion in Q2 2025, with Ethereum handling 1.4 billion queries and Arbitrum processing 1.37 billion that same quarter. Active subgraphs reached an all-time high above 15,000. The protocol supports more than 90 blockchain networks and powers data access for applications including Uniswap, Aave, Lido, and Decentraland. Grayscale added GRT to its Decentralized AI Fund in January 2026, and 37% of new Token API users are AI agents already querying blockchain data.
But the protocol faces two structural challenges. First, competition has intensified. Alchemy shut down its subgraph service in December 2025, validating the standard The Graph created, but projects migrated to faster competitors like Goldsky, Envio, and Ormi, not exclusively back to The Graph. Second, fee revenue has not kept pace with usage. Quarterly query fees reached just $98,700 in USD terms in Q4 2025, even as GRT-denominated fees rose 60.3% that quarter. The circulating market cap dropped to approximately $353 million by end of Q4 2025, down from $846 million the prior quarter.
The 2026 roadmap addresses both challenges simultaneously: expand the addressable market far beyond blockchain developers, and create new fee-generating services integrated through one decentralized protocol.
What Horizon Changed
Before Horizon, The Graph was architecturally tied to subgraphs, custom APIs that developers define to index specific blockchain events and expose them through GraphQL queries. Subgraphs work well for their purpose, but they represent one approach to one type of data access.
Horizon decoupled the protocol's core infrastructure from any single data service. Think of it like the difference between a restaurant that only serves pizza and a food hall where multiple kitchens share the same building, utilities, and payment system. Each kitchen can specialize, but they all benefit from shared infrastructure. Horizon introduced three architectural changes: a general-purpose staking protocol that extends economic security to any data service, a unified payment system that handles fees across all services, and a permissionless framework that allows new data services to integrate without rebuilding infrastructure from scratch.
This is what makes the six-product roadmap possible. Without Horizon, each new service would need its own staking mechanism, its own payment logic, and its own security model.
Six Products Targeting Six Markets
The roadmap organizes development across Q1 through Q4 2026, each product targeting a different user segment.
Subgraphs remain the foundation but are being upgraded in two directions. Cost and scaling efficiencies through a Rewards Eligibility Oracle (REO) that ties indexer rewards to actual value delivery. And AI compatibility through x402-compliant gateways with MCP and A2A integrations by Q2 2026, making blockchain data queryable through natural language in tools like Claude, ChatGPT, and Cursor, with per-query autonomous payment.
The x402 integration deserves attention. x402 is the open payment standard developed by Coinbase that activates the HTTP 402 "Payment Required" status code for native micropayments between clients and servers. It has been adopted by Stripe, Cloudflare, and Google's Agent Payments Protocol, and has processed over 100 million payment flows since mid-2025. For The Graph, this means an AI agent could query onchain liquidity data, pay a fraction of a cent in GRT per request, and receive structured results within a single HTTP request-response cycle. No account creation, no API keys, no human in the loop. In a market where McKinsey projects $3 to $5 trillion in global agentic commerce by 2030 and the agentic AI market is growing at over 44% CAGR, positioning as the default data source for autonomous agents is a high-conviction bet.
Substreams delivers high-performance blockchain data streaming for latency-sensitive applications. It has gained traction with DeFi protocols, analytics platforms, and traditional financial institutions processing real-time transactions on chains like Base, Solana, and BNB Chain. The 2026 roadmap pushes Substreams toward full Horizon integration, with a permissionless data service mainnet by Q3 and probabilistic verifiers for data integrity by Q4.
Token API provides pre-indexed, production-ready access to standardized data, token balances, prices, transfers, swaps, and NFT metadata, across 10 chains, with real-time pricing and DEX expansion planned for Q3 2026. Rather than building custom subgraphs for routine data, developers and AI agents access it out of the box.
Tycho extends into DeFi liquidity specifically. It tracks how liquidity changes across decentralized exchanges in real time, giving trading systems, solvers, and applications a single streaming interface to access prices and quotes across many DEXs. Built on Substreams, Tycho handles blockchain reorganizations automatically and works cross-chain without specialized infrastructure for each network. Private MVP ships Q1, public beta Q2, Horizon-based mainnet Q4.
Amp is the most significant new product and the clearest signal of institutional ambition. It introduces a blockchain-native SQL database built for enterprise-scale data access, audit-ready provenance, and compliance workflows. Where subgraphs serve developers and Substreams serves high-frequency applications, Amp targets regulated financial institutions that need to analyze, audit, and act on blockchain data using SQL. The roadmap cites the DTCC's Great Collateral Experiment as evidence of institutional demand. DTCC, which processes quadrillions of dollars in securities transactions annually, demonstrated in April 2025 how blockchain rails could enable real-time collateral settlement across New York, Tokyo, Paris, and London using technology from The Graph ecosystem. Amp's SQL platform and verifiable raw blockchain data are both targeted for Q4 2026.
Blockchain JSON-RPC Data Service rounds out the product layer with an experimental service exploring native read/write blockchain access, moving The Graph beyond indexed data into core RPC functionality. Research begins Q3 2026.
The Economic Flywheel
The economic argument is direct. More data services generate more protocol activity. More query volume produces more GRT fees. More fees can drive token burns. More data services require more staked GRT. As the product suite expands across developers, AI agents, analysts, solvers, and institutions, this flywheel is designed to accelerate.
Three supply-side changes are planned. Issuance will be redirected across multiple data services rather than concentrated on subgraphs alone. REO establishes a proof-of-work standard ensuring indexer rewards correlate with delivered value rather than passive staking. And Decentralized Indexer Payments (DIPs) create a mechanism for consumers and ecosystem participants to incentivize indexers directly.
Cross-chain GRT staking via Chainlink CCIP is already live on Arbitrum, Base, and Avalanche, with Solana planned for 2026. A liquid staking initiative targets centralized exchanges to improve delegation accessibility, with testnet in Q2, mainnet in Q3, and a Morpho launch in Q4.
The challenge is clear in the numbers. The Graph processed 4.97 billion queries in Q4 2025, yet generated only $98,700 in quarterly fee revenue. For comparison, GRT circulating supply has grown to 10.7 billion tokens, with approximately 3% annual inflation via indexing rewards creating roughly 343 million new tokens yearly. The roadmap's thesis is that expanding the addressable market from thousands of blockchain developers to millions of AI agents, data analysts, and institutional users will change that ratio fundamentally.
My Final Thoughts
The Graph's 2026 roadmap is a structural graduation from indexing protocol to modular data economy. The timing is aligned with three converging market forces: a blockchain infrastructure sector projected to reach $393 billion by 2030, an agentic AI economy growing at 44% CAGR toward $47 billion, and institutional blockchain adoption accelerating through initiatives like DTCC's collateral platform. The x402 integration, arriving as Stripe, Google, and Cloudflare validate the standard for autonomous machine payments, positions The Graph at a potential chokepoint in the emerging agentic data economy.
The harder question is whether these new services generate enough fee revenue to close the gap between 1.27 trillion lifetime queries and $98,700 in quarterly fees. The infrastructure thesis is sound and the market tailwinds are real. Whether The Graph captures proportional value depends on execution speed, because the competitors moving into this space are not waiting for decentralization to catch up.
Don’t forget to like and share the story!