Aethir

What it does

Aethir aggregates enterprise-grade GPUGPUGraphics Processing Unit. Originally designed to render video game graphics, GPUs turned out to be exceptionally good at the massively parallel math that AI models need. Modern AI training and inference runs almost entirely on GPUs.Like a factory with 10,000 workers doing the same simple task in parallel, versus a CPU which is more like 10 workers each doing different complex tasks. AI training involves doing simple math a million times per second on a million numbers, which is exactly what the GPU factory is designed for.GPU in the glossary → hardware (NVIDIA H100s, H200s, A100s, B200s, and B300s) from data centres, telecom providers, gaming studios, and miningProof of WorkThe original blockchain consensus mechanism where miners compete to solve computationally expensive puzzles. The winner proposes the next block and earns the rewards. Proof of Work secures Bitcoin and most pre-2020 chains.Like a lottery that runs every 10 minutes where the tickets cost electricity. Whoever spends the most electricity buying lottery tickets has the best chance of winning that round's prize. Nobody can fake the result because the proof of their work is verifiable by everyone.Proof of Work in the glossary → operations into a distributed compute network. Two products: Aethir Earth provides bare-metal GPU compute for AI trainingTrainingThe one-time process of teaching a neural network to perform a task by showing it massive amounts of example data and adjusting its internal weights until the outputs are good. Training builds the model; inference uses it.Like the years an apprentice spends learning a trade. You don't see any of the actual work, just thousands of repeated mistakes gradually becoming competence. By the end, the apprentice can do the job. The training was invisible, but the skill is now permanent.Training in the glossary → and inferenceInferenceRunning a trained AI model to produce an answer. Inference is what happens when you type a prompt into ChatGPT and get a response. The model takes your input, computes a best guess, and returns it.Like asking an expert for their opinion. The training was the decades they spent becoming an expert. The inference is the 30 seconds it takes them to answer your specific question.Inference in the glossary →. Aethir Atmosphere delivers low-latency GPU streaming for cloud gaming.

The network runs through three roles. Containers execute the actual compute workloads. Checkers verify container performance and integrity through Proof of Capacity (rewarding standby availability) and Proof of Delivery (rewarding completed tasks). Indexers match users to the best available containers based on requirements, latency, and pricing.

Founded by Daniel Wang (CEO, ex-Riot Games COO/Head of International Publishing, ex-IVC Partner) and Mark Rydon (CSO, ex-Bechtel Corporation). CTO Kyle Okamoto came from Ericsson (CEO of IoT/Automotive/Security division) and Verizon Media (Chief Network Officer). Headquartered in Singapore. The team’s background is enterprise infrastructure and gaming, not crypto-native. This shows in the product: it’s built to sell GPU compute to enterprises, not to maximise decentralisation.

That revenue-first, decentralisation-last position is exactly what Render vs Akash vs io.net sets against the permissionless and aggregator approaches.

TokenTokenA digital unit of value or access rights tracked on a blockchain. Tokens can represent ownership in a project, a right to use a service, a share of future revenue, or simply a tradable asset with no underlying claim.Like a physical poker chip a casino issues. The chip itself has no value. What makes it worth something is what it lets you do at the casino, what the casino has promised, and how much other people will pay you for it.Token in the glossary → launched 12 June 2024. The network claims 150+ enterprise clients, and its dashboard showed 433,704 GPU containers across 94 countries on 7 October 2026. In July 2025, ATHATHAll-Time High. The highest price a token has ever reached. ATH is usually quoted as a reference point for how far the current price has fallen (or risen) since the peak.Like the record lap time on a racetrack. It tells you what the car has been capable of at its absolute best, not what it will do today. Whether that record gets broken again depends on conditions that may or may not come back.ATH in the glossary → expanded to the Solana blockchain via LayerZero and Stargate bridging, enabling enterprise clients to purchase compute directly on Solana, a logical move given the concentration of AI agent builders (ai16z, elizaOS) in that space.

The product in practice

Grounded in our Aethir review, the render-vs-akash comparison, our on-chain service-fee index, and community sentiment. We haven’t run enterprise compute on Aethir, so the customer-side notes are attributed.

Who it’s for. Enterprise AI teams and cloud-gaming operators that need H100 and H200-class clusters with SLAs and support, rather than indie developers or sovereignty-minded self-hosters. Aethir sells GPU compute the way a traditional cloud company does, through Aethir Earth for bare-metal AI training and inference and Aethir Atmosphere for low-latency game streaming. The founding team comes from enterprise infrastructure and gaming rather than crypto, and the product reflects that.

What it’s like to engage with it. There are two front doors. Customers get a sales-led enterprise path with contracts, SLAs, KYC and 24/7 support, plus a thinner self-serve layer on top in the Aethir Earth rental portal and managed Kubernetes. We could not find a single independent, first-hand account of what renting from Aethir is actually like; every customer story traces back to Aethir’s own case studies, which is itself telling for a business this size. The hands-on signal that does exist independently comes from the node side, and it is cautionary.

The node reality. Aethir sold hardware to retail buyers on two tiers, and community sentiment there is mostly about disappointing economics. Checker Nodes are CPU-only uptime validators that do no GPU work, so buyers paid several hundred dollars for a device that mostly pings the network. Aethir Edge owners report payouts that swing hard on region and uptime, with community screenshots ranging from a few hundred tokens a month down to a few dollars, and a mid-stream change that redirected the Season 3 airdrop into a treasury and cut Edge emissions drew the standard reward-reduction grievance. A later Checker Node buyback, offering early holders an exit, tells its own story.

Demand, honestly. This is the one compute network in our set whose revenue reconciles on-chain: DeFiLlama indexes Aethir’s gross compute fees and our integrity check returns a match, which puts it in a different evidence class from io.net’s self-reported figures. The caveats matter, though. Protocol take is only about a fifth of that gross, as of our mid-August 2026 read the dollar run-rate had fallen for three complete quarters to near a quarter of its peak, and the scale metrics (440,000 containers and 150-plus clients as of our June 2026 read, named publishers like EA or Ubisoft) are self-reported with no independent confirmation we could find. So the revenue is the trustworthy part, and the business-scale story around it is still Aethir’s own.

The vision. Aethir’s bet is that a professionally-run GPU cloud with a token wrapper beats a permissionless marketplace on reliability and enterprise trust, and the revenue suggests the compute business has substance. The gap is the token: the code is closed-source, a 2-of-3 multisig owns the token contract, governance remains a roadmap item, and there is no fee distribution or burn, so holders capture little of that revenue. It is the sharpest version of the compute trade-off, the strongest revenue evidence in the set paired with the weakest holder claim on it.

Value proposition

Revenue. Aethir claims $127.8 million in total revenue for 2025 (January to December). The annualised run rate figures require scrutiny: Aethir’s own Q3 blog post states $166 million ARR, but a separate enterprise growth post gives $147 million ARR, which is simply Q3’s $39.8 million multiplied by four. The gap suggests the $166 million figure includes contracted recurring revenue beyond simple quarterly annualisation.

The total itself reconciles on-chain (see below), so the live question is the trajectory rather than the existence. Our index of the AETHIR_CORE contract puts the best month in the priced window at $150.2 million annualised, and the trailing ninety days at $39.6 million. Even at the lower figure this makes Aethir the highest-revenue project in the entire DeAIDeAIDecentralised AI. An umbrella term for blockchain-based projects that build AI infrastructure (compute, data, inference, models, agents) without a single central provider controlling the system.Like the difference between streaming a movie from Netflix and sharing it via BitTorrent. Netflix is fast and polished but one company controls what you can watch and what you pay. BitTorrent is messier but no single operator can shut you out.DeAI in the glossary → space by a significant margin. For context, Akash collected about $2 million of on-chain lease fees over the trailing year on DeFiLlama’s adapter, and Render does not publicly disclose USD revenue, so a comparable comparison is not possible.

The enterprise sales motion is the differentiator. Where Akash runs a permissionless marketplace and lets providers compete in reverse auctions, Aethir operates more like a traditional cloud sales organisation that happens to source GPUs from a distributed network. This is why the revenue numbers are large: enterprise contracts with committed capacity, not spot pricing.

The counter-narrative has shifted. The revenue is no longer a number you take on trust. Our own index of the AETHIR_CORE contract on Arbitrum puts the trailing year at $78.7 million of gross service fees, and DeFiLlama’s independent adapter lands within a few per cent of that on the same window and closer still on the current run-rate. Token Terminal’s separately built series agrees on recent months, and our integrity layer marks the figure a match.

What the on-chain view also exposes is the part the headline hides. The protocol’s own take is roughly a fifth of that gross, and the dollar run-rate has fallen for three complete quarters in a row, roughly halving over that stretch and landing near a quarter of the peak above. Sixty-one addresses have ever deposited a service fee, so what is falling is a handful of large enterprise contracts rather than a broad customer base thinning out.

There is a trap inside that decline, and it is the kind that flips a verdict. Denominated in ATH, the unit the contract actually emits, deposits went the other way and roughly tripled year on year. The logs settle which leg measures demand: deposit volume tracks the ATH price at a correlation of -0.90 and an elasticity of -0.48, which is what you see when prices are set in fiat and settled in the token. So the dollar leg is the demand signal, the ATH leg mostly tracks the token, and quoting either alone yields a boom or a collapse from the same 1,250 logs.

The container count of about 434,000, by contrast, is still self-certified. A dashboard exists at dashboard.aethir.com, but the underlying data comes from Aethir’s own systems. The revenue settles on-chain; the network-scale metrics don’t, and without open-source code or on-chain attestationAttestationA cryptographic proof that a piece of code is running on a specific hardware enclave in an unmodified state. Attestation lets remote users verify that a service is genuinely running what it claims to be running.Like a tamper-evident seal on a medicine bottle. The seal itself doesn't make the medicine safe, but it gives you a way to verify that nobody opened the bottle and swapped the contents before you bought it.Attestation in the glossary → they stay unverifiable.

For the sovereignty thesis, Aethir offers almost nothing. Closed source code. A 2-of-3 Safe with no timelock owns the token contract and decides who receives its remaining balance.

Governance is at least now documented: a four-phase roadmap published in late 2024 (development, community education, beta testing, full launch) with a committee comprising compute providers, validators, and ATH stakers. Full decentralised governance was targeted for Q1 2026. As of September 2026 no on-chain voting mechanism is deployed, and the two Snapshot spaces named Aethir hold only test proposals from January 2025.

Even if governance ships as promised, you’re still renting GPUs from a centralised enterprise that happens to distribute hardware across 94 countries. That’s useful. It’s just not meaningfully different from AWS with more varied suppliers underneath.

Tokenomics

ATH launched via node sale in March-May 2024, raising $148.4 million from 74,673 checker nodes sold at $440-$12,634 each. This is by far the largest node sale in DeAI. A pre-Series A round in July 2023 raised $9 million at a $150 million valuation, led by HashKey Capital with participation from Animoca Brands, Maelstrom (Arthur Hayes), Sanctor Capital, and others.

Total supply is 42 billion ATH. Distribution:

• Checker nodes & compute providers: 50%, the bulk allocation, vestingVestingA schedule that locks up tokens allocated to insiders, investors, and team members, releasing them gradually over months or years. Vesting prevents insiders from dumping on public buyers immediately after launch.Like a new employee's stock options at a startup. You don't get all the shares on day one. They unlock over four years so you stick around and do the work rather than cashing out and leaving.Vesting in the glossary → linearly over time
• Team: 12.5%, 18-month cliffCliffA waiting period at the start of a token vesting schedule during which no tokens unlock at all. After the cliff ends, tokens begin releasing according to the vesting schedule.Like a probationary period at a new job. You don't get your stock options on day one. You wait 12 months to prove you'll stick around, then everything starts unlocking normally.Cliff in the glossary → from TGETGEToken Generation Event. The moment a project's token first becomes tradeable. TGE is when vesting clocks usually start, when liquidity hits exchanges, and when public price discovery begins.Like the IPO day for a startup. Everything that happened before TGE was private valuations and paper agreements. Everything after is the public market deciding what the thing is worth in real time.TGE in the glossary → (December 2025), then 3-year linear daily vesting until December 2028
• Investors: 11.5%
• DAODAODecentralised Autonomous Organisation. A way to coordinate decisions and manage a treasury using token-weighted voting instead of a traditional company structure. Token holders propose and vote on changes directly.Like a shareholder-run company where every shareholder can vote on every decision, the votes are public, and the company can't do anything the shareholders don't approve. The coordination is messier than a normal company but nobody has unilateral control.DAO in the glossary → Treasury: 7.5%
• Ecosystem development: 7.5%
• AirdropAirdropDistributing tokens for free to eligible wallets, usually to reward early users, bootstrap a community, or decentralise token ownership away from a small group of insiders at launch.Like a supermarket handing out free samples to people who already shop there. The samples cost the supermarket nothing to print. The goal is to convert casual shoppers into loyal customers by giving them something tangible to talk about.Airdrop in the glossary →: 6%
• Advisors: 5%
• Public/private sale: 0.17%

In October 2025, the Aethir Foundation redirected 1.26 billion ATH tokens originally earmarked for the Season 3 Cloud Drop airdrop into a new Digital Asset Treasury (DAT). The DAT functions as a strategic compute reserve: it buys ATH on-market, stakes them to onboard new GPU hosts, books compute on the platform, sells that compute for dollars, and uses the proceeds to buy more ATH. The tokens cannot be distributed directly or sold on open markets. Predictive Oncology (NASDAQ: POAI) separately announced a $344 million private investment in ATH for its own DAT-based Strategic Compute Reserve, a noteworthy institutional endorsement, though the relationship between the two DAT structures is unclear.

On 24 August 2026 Aethir announced ACCELERATE: 10 sites across the US and Europe, up to 20MW in total, for NVIDIA B300 and GB300 clusters of 64 to 256 nodes. Aethir says it’s targeting up to $700 million in contracts by the end of 2026, with Axe Compute’s Axe Build unit owning and operating the sites. The same post says an updated IDC tokenomics policy “will include a burn mechanism and a variable platform fee”. No deployed contract backs that yet, so the no-burn position in the Returns section stands until one ships.

Circulating supplyCirculating SupplyThe number of tokens currently in circulation and tradeable on the open market. Differs from total supply (which includes locked or unvested tokens) and max supply (the upper limit, if there is one).Like the number of cars on the road today versus the number ever produced. Some are in showrooms, some in junkyards, some still at the factory. Only the ones on the road count toward what people are actually driving.Circulating Supply in the glossary → sits at roughly 47.93% of the 42 billion maximum, so roughly 52% of supply is still locked or vesting.

The compute provider allocation alone (21 billion tokens) will continue entering circulation for years. Team tokens started unlocking in December 2025 and vest daily until December 2028.

The gap between circulating market cap and FDVFDVFully Diluted Valuation. The market cap a token would have if every token that will ever exist were already in circulation. FDV is what the project would be worth if all locked, vesting, or unminted tokens were trading today.Like valuing a startup based on what every share would be worth if all the unvested employee options had already been exercised. The number is bigger and uglier than the official market cap, but it tells you the true ceiling.FDV in the glossary → tells you everything about the dilution ahead. ATH was at launch in June 2024. The token is down more than 90% from that peak.

StakingStakingLocking up a cryptocurrency to help secure a blockchain network, usually in exchange for rewards. The locked tokens act as a security deposit that can be taken away if the staker misbehaves.Like putting down a large rental deposit for an apartment. You get the money back if you behave, you earn interest while it's locked, and the landlord takes it if you trash the place.Staking in the glossary → exists across three pools: Gaming, AI, and EigenLayer pre-deposits. Lock-ups run up to 4 years with a 30-day withdrawal vesting period. If pool utilisation exceeds 85%, withdrawals are blocked entirely. APY is not publicly disclosed, which is a red flag. If the yields were attractive, they would be advertised.

Token contract is on Ethereum mainnet (0xbe0Ed4138121EcFC5c0E56B40517da27E6c5226B) with operations and rewards on Arbitrum. The contract is verified on Etherscan: Solidity v0.8.18, ERC-20 with ownership held by a 2-of-3 Gnosis Safe and no timelock. The contract hard-codes a 42 billion maximum supply, and totalSupply() already equals it, so the owner-only mint can add nothing and supply is fixed. The Safe’s remaining power is an owner-only whitelist that pays out the small balance the contract still holds, so two of three keyholders decide who gets those tokens.

Listed on OKX, Bybit, KuCoin, Bitget, Gate.io, MEXC, and others. Coinbase listed ATH in March 2025, adding tier-1 exchange access. ATH was also added to Binance Alpha (a pre-listing token selection pool, not a full exchange listing) with a community Vote to List campaign in April 2025. As of March 2026, ATH isn’t listed on Binance for spot trading. Daily volume is healthy relative to market cap.

How to participate

Stake ATH. Choose Gaming, AI, or EigenLayer pools. Lock-up up to 4 years, 30-day withdrawal vesting. APY undisclosed. No compounding. Pool utilisation above 85% blocks withdrawals. Requires a walletWalletSoftware that stores the private keys needed to control tokens on a blockchain. A wallet does not actually hold any tokens. The tokens live on the chain. The wallet holds the keys that prove you own them.Like the key to a safe deposit box. The key doesn't contain your valuables. The valuables sit in the bank's vault. The key is what proves you're allowed to open the box and take them.Wallet in the glossary → and some ATH. Technical skill: basic. The main risk is the undisclosed yield: you are locking tokens without knowing the return.

Run a checker node. Checkers verify container performance. Requires purchasing a node licence (node sale has concluded, secondary market via eATH tokens). Technical skill: intermediate. Hardware requirements are modest since checkers verify rather than compute.

Provide GPU compute. Supply enterprise-grade NVIDIA GPUs to the network as a Cloud Host. Technical skill: advanced. Hardware: NVIDIA H100/H200/A100/B200 minimum. Capital requirements are high, and enterprise GPU hardware is expensive. This is the participation method that generates the revenue, but the barrier to entry is substantial.

Honest assessment

Freedom Score: 30/100

Aethir scores among the lowest Freedom Scores on this site. Here is why:

Infrastructure Decentralisation: 9/20. The hardware is geographically distributed: about 434,000 containers across 94 countries is substantial coverage by Aethir’s own metrics. But all protocol logic, container management, and indexing is controlled by Aethir. The code is closed source. You cannot run an independent indexer or checker without Aethir’s permission.

Governance Decentralisation: 4/20. A four-phase governance roadmap was published in late 2024, with a committee structure comprising compute providers, validators, and ATH stakers. Governance areas include product direction, network parametersParametersThe internal numbers (weights and biases) inside a neural network that get adjusted during training. A 70-billion-parameter model has 70 billion adjustable internal numbers encoding everything it has learned.Like the synapses in a human brain. Each parameter is a tiny dial that gets nudged a little during training. With enough dials, the network can represent surprisingly complex patterns. The total parameter count is roughly how much "brain" the model has.Parameters in the glossary →, token economics, and staking rewards. Full launch was targeted for Q1 2026, but as of September 2026 no on-chain voting mechanism is deployed, and the only Snapshot proposals are tests from January 2025. The DAO Treasury holds 7.5% of supply. Score remains at 4 until governance is demonstrably operational, not just roadmapped.

Token Distribution Fairness: 5/15. 50% to compute providers is structurally positive but 29% goes to team, investors, and advisors. The node sale raised $148 million from an unknown number of unique buyers, so concentration risk is unassessable. Public sale was 0.17%.

Censorship Resistance: 5/15. Closed source code means Aethir controls what runs on the network. The token itself has no pause, freeze or blacklist function, and its owner multisig can’t restrict holders’ transfers. Enterprise focus suggests compliance orientation, not censorship resistance.

Data Sovereignty: 5/15. Enterprise workloads run on distributed containers, but Aethir controls the matching and indexing layer. No evidence of encryption guarantees or privacy-preserving computation.

Open Source Transparency: 2/15. Seven own public repositories across two GitHub orgs, and not one of them carries a licence, so a reader has no right to use, modify or redistribute any of it. What’s published is the node-operator edge (the checker client, a host agent, a JS SDK); the container orchestration, checker logic and indexer code are absent from both orgs. The token contract is verified on Etherscan, which is the bare minimum. You can’t independently verify Aethir’s network-scale and performance claims from the code, even though the service-fee revenue does settle on-chain.

Returns Score: 62/100

Token Utility: 12/20. ATH is used to pay for compute, stake in pools, and operate checker nodes. Real utility exists: compute providers are paid in ATH. But staking yields are undisclosed, governance is nominal, and there is no access gating for general users. The token is useful but not essential to most participants.

Value Accrual: 14/20. Compute providers earn ATH for services. Staking pools direct tokens toward AI and gaming workloads. But there is no fee distribution to token holders, no buy-and-burn mechanism, and no revenue sharing. Value accrual is indirect: token demand comes from compute payment and staking requirements, not from a mechanical link between protocol revenue and token value.

Supply Dynamics: 8/20. 42 billion max supply with only 47.93% circulating. 50% allocated to compute providers vesting over years represents sustained dilution pressure. Team tokens vesting until December 2028. No burnBurnPermanently removing tokens from circulation by sending them to an address that no one controls. Burns reduce total supply, which (all else equal) makes each remaining token worth more of the network's value.Like a company buying back its own shares and shredding them. The company's total value stays the same, but each remaining share now represents a slightly bigger slice of that value.Burn in the glossary → mechanism. The FDV-to-market-cap ratio is well above 1x, quantifying the dilution ahead. Score capped low given the combination of heavy remaining emissionsEmissionsNew tokens created and distributed by a blockchain protocol over time as rewards to validators, stakers, or miners. Emissions fund network security and participation at the cost of diluting existing holders.Like a company that pays employees partly in newly printed shares. Every year the total number of shares goes up, which means existing shareholders own a slightly smaller slice of the same company unless the company grows faster than the printing.Emissions in the glossary → and no offsetting burns.

Revenue Sustainability: 16/25. Aethir’s 2025 revenue headline reconciles on-chain: our own AETHIR_CORE index puts the trailing year at $78.7 million of gross service fees, DeFiLlama’s independent adapter agrees to within a few per cent, and our integrity layer marks it a match. That makes for demonstrable product-market fit at a scale no other DeAI project approaches. Three caveats keep it short of the top band. The protocol keeps only about a fifth of that gross; the DAT booking compute on Aethir’s own platform leaves an open question about how much of the volume is self-referential; and the dollar run-rate has fallen for three complete quarters, from a best priced month worth $150.2 million annualised to $39.6 million over the trailing ninety days. The rubric’s 17-21 band asks for revenue growing quarter on quarter, and this no longer does, so the September 2026 review moved the score from 18 to 16. That is the top of the band below, which keeps the credit for verified scale while removing the credit for a trajectory that has reversed.

LiquidityLiquidityHow easily a token can be bought or sold without moving the price. High liquidity means you can enter or exit large positions quickly at the quoted price. Low liquidity means even small trades can swing the market.Like the difference between selling a house and selling a share of Apple stock. The house might be worth more on paper, but finding a buyer at that price takes weeks. The Apple share converts to cash in one click.Liquidity in the glossary → & Access: 12/15. Listed on OKX, Bybit, KuCoin, Bitget, Gate.io, Coinbase (March 2025), and several others. Binance Alpha pre-listing pool inclusion (not a full spot listing) with a Vote to List campaign in April 2025. As of March 2026, not listed on Binance for spot trading. Daily volume provides reasonable entry and exit for most position sizes. The Coinbase addition brings genuine tier-1 access that was previously absent. Score raised from 11 to reflect Coinbase listing.

Quadrant: C (Low Freedom, Moderate Returns)

Aethir sits firmly in Quadrant C: potentially profitable but centralised. The revenue story is compelling and now reconciles on-chain. The freedom story is improving (governance is finally being documented) but remains weak. This is an enterprise GPU business with a token, and governance window-dressing doesn’t change the fundamental architecture.

Recent security incident

On 10 April 2026, the AethirOFTAdapter contract on BNB Chain was exploited via an unauthorised transferOwnership call. PeckShield’s initial estimate put gross drainage at roughly 423,000 ATH (about $400,000), with attacker funds bridged from BNB Chain to Tron via Symbiosis Finance. Aethir’s monitoring caught the activity within minutes, engineers disconnected compromised contracts, and the main ATH supply on Ethereum was unaffected. End-user losses were capped at under $90,000 according to Bankless Times reporting on the incident. Aethir partnered with Binance, Upbit, Bithumb and HTX to blacklist attacker wallets and engaged ZeroShadow on attack-path analysis, with a compensation plan and detailed postmortem promised within a week. On 15 April 2026 Aethir’s official account said the full ATH compensation plan was live with a Discord claims process and the Uniswap LP fully restored; whether every individual claim has since been paid is Discord-gated and not independently verifiable. We’re not aware of any prior publicly documented protocol-side security incident affecting ATH user funds, but we haven’t done an exhaustive historical review.

Key risks

• Revenue composition and trajectory. The headline revenue reconciles on-chain, so the open risk has moved from fabrication to make-up and direction. The protocol keeps only ~20% of gross, the run-rate has fallen from its mid-2025 peak, and the DAT booking compute on Aethir’s own platform with redirected airdrop tokens raises how much of the volume is durable external demand rather than self-referential.
• Related-party buildout partner. Axe Compute Inc. (Nasdaq: AGPU), the partner named for ACCELERATE, is the company that filed with the SEC as Predictive Oncology until December 2025 (EDGAR CIK 0001446159), which makes it the ATH treasury company described in Tokenomics. Aethir’s own 17 September 2026 post discloses that the Aethir Foundation is Axe Compute’s largest shareholder, through its 2025 treasury transaction. The same post says Axe sold its Helomics oncology subsidiary on 15 September, leaving GPU compute as its whole business. Contracts routed through a partner that holds ATH as its treasury and whose largest shareholder is the Foundation need reading as related-party business.
• Closed source. Every technical claim (node count, compute delivered, container performance) is unverifiable. This is unusual for a project claiming decentralisation.
• Supply overhang. Roughly 52% of tokens are locked/vesting. Compute provider emissions (50% of supply) will continue for years. Team/investor unlocks run until December 2028. The DAT’s 1.26B redirected tokens add another mechanism affecting circulating supply.
• Admin multisig. Supply is capped at 42 billion in code and fully minted, but a 2-of-3 Safe with no timelock controls the whitelist that distributes the tokens still held by the contract. Most serious DeFiDeFiDecentralised Finance. Financial services like lending, trading, and yield farming built on smart contracts instead of traditional banks or brokerages. DeFi protocols are usually permissionless and global.Like a vending machine that can give you a loan, swap your currencies, or invest your savings. Nobody is behind the counter, the rules are written into the machine itself, and anyone with money in the right format can use it.DeFi in the glossary → projects have moved that kind of control behind a timelock or a vote.
• Governance uncertainty. Full decentralised governance was targeted for Q1 2026, but no on-chain voting mechanism was deployed as of September 2026. The governance committee structure is documented but untested.
• Enterprise concentration. 150+ clients sounds healthy, but if a small number of large contracts represent the majority of revenue, client departure risk is real.

Score change log

Date	Score	Change	Reason
2026-10-07	Data	N/A	Container count refreshed from 440,000+ to about 434,000 (Aethir’s dashboard, 7 October 2026), still self-reported. Scores unchanged.
2026-10-04	Editorial	N/A	Product-in-practice sweep: anchored the revenue decline to our mid-August 2026 fee-index read, dated the container and client self-reports, and softened ‘best-run business’ to ‘strongest revenue evidence’. Scores unchanged.
2026-10-04	Data	N/A	Related-party risk tightened: Aethir’s 17 September 2026 post discloses the Foundation is Axe Compute’s largest shareholder, not just a shareholder; Axe sold Helomics on 15 September. Scores unchanged.
2026-09-25	Data	N/A	Token contract corrected: supply is capped at 42B in code and fully minted, so the multisig can’t mint; it does hold an owner-only whitelist over the contract’s remaining balance, which the 24 Sep row missed. Scores unchanged pending review.
2026-09-24	Data	N/A	Token contract corrected to the integrity finding (mint and owner only, no whitelist); governance restated as no on-chain voting as of September 2026; container and client counts attributed to Aethir; ATH drawdown and holder counts de-hardcoded; Akash comparison updated to its trailing-year lease fees. Scores unchanged.
2026-09-24	Data	N/A	Added ACCELERATE (24 Aug 2026) and its announced IDC tokenomics policy, whose burn and variable fee are stated intent with no deployed contract. Added a related-party risk: partner Axe Compute is the renamed Predictive Oncology (EDGAR), and the Aethir Foundation holds shares in it. Scores unchanged.
2026-09-23	Data	N/A	Integrity records re-cut: 8 claims now quote Aethir’s own pages verbatim and 10 are our own measurements. Scores unchanged.
2026-09-17	Data	N/A	Corrected the open-source evidence: the inflated verdict recorded “zero public GitHub repositories”, read off github.com/Aethir, a personal account. Aethir’s own docs name AethirApp, and it plus AethirCloud carry seven own repos, all unlicensed; verdict stands, 2/15 flagged for monthly review.
2026-09-06	Data	N/A	Research JSON resynced to the frontmatter Returns breakdown, which is the source. Revenue Sustainability 17 and Liquidity & Access 11 in the JSON were pre-2026-06 and pre-2026-03 values that two logged moves never propagated. No published score changed.
2026-09-05	Returns	Revenue Sustainability 18 → 16 (Returns 64 → 62, C held)	Executes the decision of 2026-08-17. The 17-21 band requires revenue growing quarter on quarter; the USD leg has fallen across three complete quarters to about a quarter of peak, so the score moves to the top of the band below, keeping credit for verified scale.
2026-08-17	Data	N/A	Revenue figures re-sourced and bound live: the ”~$163.5M all-time” headline came from a hand-run read the committed index cannot reproduce, so all-time dollar claims are dropped and the trailing year is bound via <Fig>. Trajectory restated from “fallen since” to the measured three-quarter decline. Score held, flagged for September.
2026-08-17	Data	N/A	Service-fee index committed and backfilled (1,250 DepositServiceFee logs). Two corrections: deposits denominated in ATH ROSE 2.33B to 7.43B year on year while the USD conversion fell, and the “under $1M by June 2026” monthly figure was overstated (indexed: $3.6M). Demand grade held, flagged for September.
2026-08-15	Data	N/A	Circulating supply de-hardcoded: the absolute token count moved out of prose and the TokenFactsStrip into the live <CirculatingSupply /> component per the circulating-supply convention. Percentage unchanged at 47.93%. No score change.
2026-07-06	Editorial	N/A	Added a “The product in practice” section (per specs/product-comparison-framework.md): enterprise sales-led, revenue reconciles on-chain but scale metrics are self-reported, and node economics are the main independent gripe. No score change.
2026-07-01	Data	N/A	July monthly review resolves two deferred Aethir items, no score change. (1) Bridge-exploit compensation (10 Apr 2026 AethirOFTAdapter, ~$400K gross / under $90K user loss): Aethir’s own account confirmed the full ATH compensation plan live with a Discord claims process and the Uniswap LP fully restored as of 15 Apr 2026; individual payout completion is Discord-gated and not independently verifiable, so the body does not assert “all users reimbursed”. (2) eATH (EigenLayer AVS) redemption window opened 24 June 2026 (slipped from the originally circulated 13 June), 15-day unbonding period, ~7.5% pre-depositor reward on a 90-day claim window. Freedom hold stands: the exploit impact was already priced in.
2026-06-10	Returns	63 → 64	Revenue Sustainability 17 → 18. Integrity layer reconciled Aethir’s revenue on-chain (DeFiLlama ~$167.9M gross all-time, Token Terminal, our AETHIR_CORE index ~$163.5M; calendar-2025 ~$117.8M; verdict match), correcting the prior “self-reported / not independently verified” framing across the body, snapshot, pillars, cons and Returns verdict. Score nudged +1: independent verification confirmed, held back by a falling on-chain run-rate (peaked ~$145M annualised mid-2025; ARR verdict outdated) and self-referential-composition questions (DAT). Doesn’t cross a grade boundary (C).
2026-06-02	Data	N/A	Circulating supply corrected from 41.6% (~17.5B) to 47.93% (~20.1B of 42B) per CoinGecko; “locked/vesting” updated 58.4% → ~52% across body, TokenFactsStrip, cons and Supply Dynamics evidence. Score unchanged (dimension already prices in heavy remaining emissions). June monthly review.
2026-05-02	Data	N/A	AethirOFTAdapter bridge exploit on BNB Chain (10 April 2026) added to risks. Gross ~$400K drained, user losses capped under $90K. Postmortem and compensation plan pending. No score change applied yet pending postmortem.
2026-03-12	Returns	62 → 63	Coinbase listing, DAT treasury mechanism. Liquidity improved.
2025-03-03	Both	N/A	Initial publish. Freedom 35/100, Returns 62/100.