A New Service Standard for Subgraph Indexing Rewards
Last week, The Graph Foundation laid out : to move from coordinating external development to directly building, maintaining, and scaling The Graph Network. Improving the network's economics is one of its stated priorities.
Three initiatives will carry that work forward in the coming months. Each ties what Indexers earn more closely to the value they deliver, and what Delegators earn follows from that. The Rewards Eligibility Oracle (or REO) ties indexing rewards to the service delivered by Indexers on The Graph Network. The Liquid Staking Initiative will make staked GRT usable while it stays staked. Direct Indexer Payments (or DIPs) will give both The Graph Foundation and other Subgraph Gateways a way to pay Indexers directly for the service levels the network needs. REO is running now, the Liquid Staking Initiative is opening its first phase this week, and Direct Indexer Payments is slated for release later this year.
This post is about REO.
Why Add a Rewards Eligibility Oracle?
Indexing rewards on The Graph Network have been distributed according to curation signal and allocated stake. Neither of those measures whether an Indexer served a single query, which means that rewards are not directly tied to the value that Indexers provide to The Graph’s consumers.
In 2025 alone, 15.2% of indexing rewards went to Indexers that were not delivering value to the network. Those Indexers were not serving queries, and they were not indexing data that was of direct use to consumers. REO changes the condition for receiving those rewards. Introduced in and deployed as part of , it ties eligibility to service delivered to end consumers. The mandate describes the shift in one line: indexing rewards move from proof of presence to proof of work. Indexers that meet a published minimum standard stay eligible. Indexers that fall below it stop receiving indexing rewards until they meet it again.
Total issuance does not change. Who receives that issuance does.
What the Rewards Eligibility Oracle Checks
The standard measures one thing today: whether an Indexer is serving queries.
Eligibility comes from serving at least one qualifying query on five or more separate days in a rolling 28-day window. A qualifying query has to satisfy three conditions at the same time: The query must return an HTTP 200 (successful) status. It must respond with that valid response in under 5,000 milliseconds. Lastly, the Indexer serving the query must be within 50,000 blocks of chainhead.
REO evaluates service data once a day and renews eligibility on-chain for the Indexers that qualify. Eligibility lasts 14 days from renewal, so an Indexer who keeps serving is re-qualified long before the previous period runs out.
Anyone can see the results. The publishes the status of every Indexer on the network, along with how many consecutive days each has qualified and the transaction that last renewed them.
As of this writing, the dashboard lists 97 Indexers: 49 currently eligible, one in a grace period, and 47 that have not yet qualified. This is likely to change as Indexers recognize the impact and make changes to re-qualify or move on from their indexing responsibilities.
A count of operators and a share of rewards measure different things, so the two should not be read against each other. What the dashboard shows is the standard running on live data, in public.
Why the REO Threshold Starts Here
Five days out of 28 is only about 18% uptime, which is a deliberately low bar.
The first version of a standard should catch only the cases nobody disputes. An Indexer serving real traffic clears five days without noticing the requirement exists. An Indexer serving nothing at all does not. Between those two, the threshold leaves room for maintenance windows, migrations, temporary infrastructure problems, and smaller operators who are just ramping up.
A modest economic floor sits underneath it. Queries only count toward eligibility on Subgraphs carrying at least 500 GRT in curation signal, which puts a small cost on manufacturing traffic against a self-published Subgraph. That floor matters more as the serving requirements grow.
While the threshold starts low, the criteria will get stricter over time. Requirements around uptime, reliability, and the breadth of Subgraphs an Indexer supports are all under consideration. None of those specifications are in effect today, and none of them will arrive without notice. The Graph’s quality of service should lead the industry and the REO will be one of the tools that can be leveraged to deliver the ecosystem to that higher standard.
What Happens When an Indexer Does Not Qualify?
An Indexer that does not meet the standard stops receiving indexing rewards. That is the point of the change. Rewards are withheld, not confiscated. When an ineligible Indexer presents a Proof of Indexing that would otherwise pay, the transaction reverts. The rewards stay pending. If that Indexer regains eligibility before the allocation goes stale, the rewards are still there to claim.
Nothing happens to an Indexer's stake. They can keep operating, keep serving queries, and keep earning query fees the entire time. Eligibility returns as soon as they meet the standard again, with no application and no appeal process.
Indexers are also not stuck. While ineligible, an Indexer can present a “Zero Proof of Indexing” transaction to reset the staleness clock and keep an allocation alive, or close the allocation and exit cleanly.
Indexers can also check why a given Proof of Indexing paid or did not. Every presentation records the condition that applied, readable from the network Subgraph:
{allocations {idlatestPoiCondition}}
None of that was visible before this upgrade.
What REO Means for Delegators
Delegated stake follows the Indexer that holds it. If an Indexer is not eligible, their Delegators do not receive indexing rewards either. This is the part of the change with the widest reach. Choosing an Indexer has always been the Delegator's decision. That decision now determines whether a delegation earns indexing rewards at all.
Delegators can see it coming. Every Indexer's eligibility status is , updated daily, along with their qualifying streak. Delegators can check the Indexer they are staked to at any time to monitor when an Indexer's status changes. The status to watch for is "grace," which means an Indexer is still eligible but needs to renew soon.
A Delegator whose Indexer stops qualifying will need to move that delegation to keep receiving indexing rewards. The lists active Indexers alongside what they have allocated and the Subgraphs they serve. Of course, moving delegated stake does take time. Undelegating is subject to a thawing period of 28 epochs before the GRT can be moved, which is why watching an Indexer's status beats reacting after it lapses.
A second route is on the way. The Foundation is opening the first phase of the GRT Liquid Staking Initiative, which turns staked GRT into a liquid token called stGRT, representing the holder's staked position and accrued rewards. The vault was deployed by Avantgarde Finance using Enzyme vault tooling and is managed by the Foundation.
Phase 1 of Liquid Staking is a soft launch that is deliberately small. The launch is open to a limited group of existing holders while the accounting and deposit cycles are validated. A wider rollout for Delegators will follow in the coming weeks. The community can learn more about the first stage of liquid staking via the public , and supporters who want to be considered for early access can email .
Note: This section is informational only. It is not financial, investment, legal, or tax advice, and nothing here guarantees value, yield, or returns. Rewards are variable and depend on network conditions. Features and timelines are subject to protocol governance and may change.
Why Trust REO?
An off-chain oracle deciding who gets paid deserves scrutiny.
The criteria were published in advance of REO going live. They live in a public document with three sections: the requirements in force today, the changes that are coming, and a changelog of everything that has changed before, each with a written justification. Announced changes carry roughly a 14-day window before they take effect, and an Indexer who has already qualified keeps claiming through their full period even if the criteria tighten during it.
The code is open, so anyone can read the exact logic that decides eligibility rather than taking a description of it on faith. The oracle runs on a fixed daily schedule and writes a record of each evaluation, showing which Indexers qualified and why.
REO also takes a conservative approach to potential downtime. If the oracle stops posting updates for seven days, every Indexer is treated as eligible automatically. A broken oracle cannot halt rewards for the network, and governance can disconnect it entirely at any time, which restores the previous behavior immediately. That is what makes the standard safe to tighten later: a mechanism that cannot accidentally deny the whole network can be adjusted without much risk.
GIP-0079 was authored in October 2023 and ratified by The Graph Council. The time in between went into the measurement infrastructure, the dashboard, and the fail-safes that make enforcing a standard fair.
What Else is New?
REO arrives alongside two pieces of protocol plumbing that do not change anything today.
The Issuance Allocator lets protocol issuance be directed to more than one destination, instead of flowing entirely to indexing rewards. At launch it routes 100% to indexing rewards, exactly as before, so the rate does not change.
The contracts behind Direct Indexer Payments (DIPs) are also deployed, but they are dormant for now. The DIPs contracts are the on-chain rails for the second of the three initiatives: agreements that let the Foundation and Subgraph Gateways pay an Indexer to index specific data, rather than relying on curation signal to approximate demand. This first release is a tool for the Foundation and Gateway Operators rather than something data consumers use directly. When those contracts are activated, a change to the indexing rewards rate will be communicated at that time.
Both exist so that issuance can respond to what the network is being asked to do. Neither is doing anything yet. The oracle is the part that is live today.
Bottom Line: What REO Changes
The Graph Network is an open protocol connecting the people who need blockchain data with the operators who serve it. Until now, the reward side of that marketplace paid for presence. It now pays for service.
Indexers that do not meet the standard are expected to either improve their service or move to another role in the protocol, and some may choose to participate as Delegators instead. The protocol issues the same amount of GRT it issued before. More of it now reaches Indexers who are serving queries.
That is a more economically efficient protocol. It is also the first of three big moves that The Graph community can look forward to for changing how the next iteration of the protocol’s economics will work.
About The Graph
The Graph is a suite of blockchain data infrastructure products that extract, process, and deliver scalable blockchain data solutions across 60+ networks. The Graph enables application developers, data analysts, AI agents, and enterprise teams that need structured, real-time access to blockchain data. Products include Subgraphs, Firehose, Substreams, and Amp. As of early 2026, The Graph has served over 1.27 trillion queries to more than 75,000 projects, powered by a network of independent Indexers around the world.
Follow The Graph on , , , and . Join the community on The Graph’s , join technical discussions on The Graph’s .
