The Graph indexed 11.6 billion queries in the last six months and the market paid roughly no attention. Price up 1.7% over seven days while the crypto tape lifted 4.3%. The panda has seen this before: a piece of infrastructure that everyone uses and nobody owns.
Where GRT sits in mid-2026
The numbers are blunt. According to CoinGecko's GRT page, GRT trades around $0.0198 with a $213.9 million market cap on 10.80 billion circulating tokens, against a max supply of 10.80 billion. Fully diluted valuation: also $213.9 million. The unlocks story is done. There is no inflation hangover, no team cliff coming next quarter, no narrative escape valve. What you see is what you trade.
That makes GRT a strange creature in 2026. Most "infrastructure" altcoins still carry FDV-to-mcap multiples of 2x to 6x, which is a polite way of saying the market has not been paid yet. The Graph essentially has been. Either the asset works at this size or it does not.
How does The Graph actually make money?
Worth answering directly, because the headline numbers are humbling. According to DefiLlama's protocol page for The Graph, the network generated about $617,720 in annualised fees as of June 2026. Trailing 30-day fees: $35,387. Cumulative fees since inception: $1.35 million.
Set that against the $213 million market cap and the price-to-fees ratio is roughly 350x. By the standards of public SaaS that would be obscene. By the standards of crypto infrastructure tokens, it is roughly mid-table, and the comparison set is itself suspect.
The economic model is simple in shape, painful in scale. Data consumers pay query fees in GRT to indexers, who stake GRT and serve curated subgraphs. Curators stake on subgraphs they expect to be queried. Delegators stake on indexers without running infrastructure. Three sinks, one issuance line, all gated by query demand. The panda raises an eyebrow at the level, not at the design.
Tokenomics: the rare nearly-finished altcoin
Issuance is the cleanest part of the GRT story. The token launched in December 2020 with a 10 billion initial supply and a 3% annual indexing-reward issuance, partially offset over time by query-fee burns and governance parameter changes. As of June 2026, circulating supply sits at 99.996% of max supply per CoinGecko's published figures above.
Practically, GRT holders no longer get diluted at the structural level. Future supply changes route through governance, which voted in 2024 to retire the centralised hosted service and direct all indexing rewards to the decentralised network on Arbitrum. According to Blockchain.News on The Graph's Arbitrum migration, the protocol shifted 100% of indexing rewards to L2 on June 28, 2024. Validators, infrastructure and economics now live on Arbitrum, not Ethereum mainnet.
Treasury sits with The Graph Foundation, which has historically funded core teams (Edge & Node, StreamingFast, Semiotic Labs) through grants rather than token sales. No surprise OTC drops, no quarterly "ecosystem unlock" press release. For an altcoin in 2026, the absence of an unlock schedule worth fearing is a feature, not a bug. Most of the cluster cannot say the same.
Competitive position: subgraphs everyone forgets they use
The Graph's product is subgraphs: defined data slices of a chain, indexed continuously, queried via GraphQL. The competitive set in 2026 has narrowed to three serious peers.
- Goldsky: VC-backed hosted indexer, raised on the "managed The Graph" pitch, popular with teams that want zero operational headache.
- Subsquid (now SQD): open-source indexing with its own token model, focused on multi-chain and high-throughput cases.
- Chainstack Subgraphs: enterprise-flavoured managed indexing, RPC-first customer base.
The honest read: hosted competitors took the easy half of the market, the developers who do not want to think about indexers, curators, delegators or GRT mechanics. The decentralised network kept the other half, teams that want censorship-resistant data access or refuse to depend on a single hosted vendor's pricing page.
According to a Messari report on State of The Graph Q2 2025, the network shipped 1,673 new subgraphs in Q2 2025, a 46.3% quarter-on-quarter jump, with active indexers up 2.8% and allocated stake up 5.3%. That was the first quarter of indexer growth after nine consecutive quarterly declines. Not a reversal yet. The first dent in one.
The Graph also indexes Solana, Base, Optimism, Polygon and Avalanche, which removes the "Ethereum-only museum piece" charge but does not yet show up in fee revenue.
What to watch next: Horizon and the AI agents bet
Two catalysts sit ahead.
First, the Horizon Subgraph Service mainnet, scheduled for Q1 2026 per The Graph Foundation roadmap. Horizon turns the protocol into a more modular data layer, with independently parameterised services on top of a shared staking base. The bet: indexers can serve other paid products (substreams, firehose, MCP-style endpoints) without the protocol spinning up new tokens. If it works, the fee line stops being a single-product story.
Second, the AI agent thesis. Agents need structured on-chain data, fast, in formats LLMs can consume. The Graph's natural-language query endpoint, rolled out earlier in 2026, is an early bet on that surface. Whether it captures revenue or gets commoditised by RPC providers selling cached JSON is the live question.
In the wider altcoin context, GRT sits in an oddly quiet corner. The altcoins cluster pillar is full of L1 and L2 tokens fighting for narrative oxygen. Tokens like Polkadot's quiet reset in 2026 and the Chainlink 2026 infrastructure thesis sit closer to GRT than to any single-chain L1: infrastructure that the wider market only remembers when it breaks. For BSC teams and memecoin trackers that consume data via off-the-shelf APIs, Dadacoin included, GRT is probably the indexer most of them already use without thinking.
The panda watches. Judgment reserved. Almost-fully-diluted, fee-light, catalyst-loaded, narrative-dead. Either Horizon ships and changes the slope, or GRT stays the most-used altcoin nobody talks about.
