Olas

What it does

Olas (formerly Autonolas) is a protocol for building and running autonomous AI agents on-chain. You build an agent, register it as an NFTNFTNon-Fungible Token. A unique blockchain-tracked asset where each token is distinguishable from every other. Where regular tokens are interchangeable, NFTs represent unique items like art, collectibles, in-game assets, or domain names.Like the difference between a $20 note and a signed first-edition novel. The notes are interchangeable, any $20 buys the same thing as any other. The book is one of a kind, and its value depends entirely on which specific book it is.NFT in the glossary → on the Olas registry, and deploy it to do useful work. If other people use your agent code, you earn rewards. If you stake OLAS and run an agent, you earn rewards based on what that agent actually does. The protocol calls this Proof-of-Active-Agent.

The architecture has three layers. The on-chain registry is an NFT-based system for components, agents, and services, enabling composable software reuse and tracking who built what. The tokenomics layer coordinates incentives through bonding (protocol-owned 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 → modelled on Olympus 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 →), developer 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 → (code top-ups for used components), and Proof-of-Active-Agent staking 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 →. The governance layer uses veOLAS (vote-escrowed OLAS, locked for up to four years) for Snapshot and on-chain voting.

Where it gets interesting is the execution layer. Agent services run off-chain as multi-agent systems using Tendermint for consensus between agent instances, while being crypto-economically secured on public blockchains via Safe multisig wallets. Deployment spans nine chains: Ethereum, Gnosis Chain, Polygon, Solana, Arbitrum, Optimism, Base, Celo, and Mode, plus Robinhood Chain since September 2026.

Three products drive usage. Pearl is a consumer app available on macOS, Linux and Windows that lets non-technical users own and operate autonomous agents with one-click staking. The Mech Marketplace enables agent-to-agent transactions, where agents hire other agents for specialised tasks. Olas Predict runs autonomous predictionInferenceRunning 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 → market trading agents on Gnosis Chain and Polymarket.

The project was built by Valory AG, a Zug-based company co-founded in mid-2021 by David Minarsch (PhD in Applied Game Theory from Cambridge, former Head of Multi-Agent Systems at Fetch.ai where he created the original AEA stack), David Galindo (former Associate Professor of Computer Security at University of Birmingham, 2,625 academic citations), and Oaksprout the Tan (pseudonymous 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 → community contributor). Team size is approximately 30. Ellie Hirschhorn (ex-CoinDesk COO, Harvard MBA) is Chief Growth Officer.

Funding totals $18.35 million: a $4 million seed round (October 2022, led by True Ventures), a $547,580 Liquidity Bootstrapping Pool on Fjord Foundry (July 2023, 149 holders), and a $13.8 million OTC round (February 2025, led by 1kx). Other backers include Zee Prime Capital, Borderless Capital, and Signature Ventures.

Value proposition

The pitch is “co-own AI.” Build agents, register them on-chain, earn rewards when they do useful work. The protocol’s Proof-of-Active-Agent staking rewards actual agent activity, not passive 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 → lockup. This is a meaningful distinction. Most staking mechanisms reward you for doing nothing. Olas rewards you for running agents that meet performance targets.

On-chain usage is measurable in part. Olas reports 20.9 million total agent transactions, 14.8 million of them agent-to-agent through the Mech Marketplace, and 449 daily active agents on a seven-day average. That daily figure is down from the 600-plus Olas published earlier in 2026. These counters are Olas’s own, and our registry reads count a different population, so they can’t test them.

Two other figures are Olas’s own and we don’t lean on them. A share of Gnosis Chain Safe transactions circulates as a Gnosis co-founder’s quote rather than a published Olas series, so we cite no number for it. The Olas Predict agents are claimed to achieve 79% prediction accuracy, which comes from Olas’s own reporting.

Robinhood Chain joined the Olas fleet in September 2026. Olas DAO proposal 16 adopted an instance already deployed there as canonical, and the GovernorOLAS contract on Ethereum shows it executed, so the L1 Dispenser now routes staking incentives to chain 4663. On Robinhood Chain the contracts answer to the L1 Timelock, but usage is negligible so far. When we read them on 24 September 2026, the Mech Marketplace had logged 17 requests and the service registry held 9 services.

Pearl makes agent ownership accessible. Install the app, stake OLAS, launch an agent. Agents like Polystrat (Polymarket trading), Omenstrat (Omen prediction markets), and BabyDegen (DeFi) run autonomously and earn rewards based on performance. The “own, don’t rent” philosophy is aligned with the sovereignty thesis.

But here’s the problem. The marketplace turnover is $110.1K in total, by Olas’s own counter. Not per day. Total. The agent economy generates impressive transaction volume but negligible economic value. Reports of 138% staking APY are meaningless when the token has declined 99.6% from its all-time high. You can earn high nominal yields on an asset that has lost nearly all its value.

The Fetch.ai connection is notable. Minarsch built the original Autonomous Economic Agent toolkit at Fetch.ai, then left and forked it as Open AEA, removing Fetch.ai’s centralised registry dependencies. Olas is, in a meaningful sense, the decentralised successor to what Minarsch originally built at Fetch.ai. Whether that is an advantage (deep domain expertise, years of iteration) or a limitation (same fundamental approach, different token) depends on your perspective.

Tokenomics

OLAS tokenomics in brief: the token is needed to stake agents for Proof-of-Active-Agent rewards, to buy at a discount through bonding, and to lock as veOLAS for governance votes, and the staking and bonding rewards are paid from emissions. Nothing returns protocol revenue to holders today: the Mech Marketplace fee burn and the buyback-and-burn are deployed but have never burned a token, and the ETH fee switch exists only in a governance proposal (AIP-6). The one supply cut on record is a January 2025 DAO action that took about 77.9 million OLAS out of circulation for good, matching Olas’s own figure. Dilution comes from the emission curve, which had issued 52.9% of the ten-year 1 billion cap by September 2026, and from a 42.65% insider allocation whose enforced buOLAS release ended in July 2026.

OLAS is an ERC-20 on Ethereum. Maximum supply: 1 billion over the first 10 years via an S-shaped emission curve, then capped at 2% annual inflationInflationThe annual rate at which new tokens are created and added to the circulating supply. Most networks use inflation to pay validators, stakers, and infrastructure providers from freshly minted tokens rather than real revenue.Like a landlord who raises the rent every year. If your salary goes up at the same rate, you break even. If it doesn't, you get poorer without noticing, because the number on your payslip hasn't changed but the ground under it has shifted.Inflation in the glossary →. Total supply on-chain was 528.8 million in September 2026, 52.9% of the ten-year cap. CoinGecko counts a lower total of 473.7 million and puts 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 → at 299.5 million, or 63.22% of its total.

Distribution:

• Developer Incentives and Bonders: 47.35% (distributed autonomously by the protocol over 10 years via bonding discounts and developer code top-ups)
• DAO Founding Members: 32.65% (most of it claimed in 2022 into veOLAS locks of one to four years or into the buOLAS smart contractSmart ContractA program stored on a blockchain that runs automatically when its conditions are met. Smart contracts are how blockchains do anything beyond just transferring tokens — DeFi, NFTs, DAOs, and DeAI infrastructure all run on smart contracts.Like a vending machine. You put in the right input and it produces the expected output, no human operator required. The rules are fixed in the machine itself, anyone can use it, and nobody can stop a transaction in the middle.Smart Contract in the glossary →’s yearly release, all ended by July 2026)
• Valory AG (Core Contributor): 10.0% (no locking or vesting requirements)
• DAO Treasury: 10.0% (controlled by DAO governance; mostly untouched)

The insider allocation is 42.65% (founding members plus Valory), all 426.5 million OLAS of it minted on 30 June 2022 to two addresses. A Sale contract received 292.9 million and could only pay out into locks: 181.0 million went into veOLAS locksVestingA 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 → of one to four years and 111.9 million into buOLAS’s yearly release, so about 69% of the insider allocation was contract-locked from the start. The other 133.6 million went to a 2-of-3 Safe with no lock, which later put about 59 million into veOLAS locks, a vesting contract and buOLAS.

Every one of those locks had ended by July 2026, so the constraint is now history. The insider OLAS still locked today is the 52.7 million revoked inside buOLAS for good plus about 21 million in veOLAS locks three early holders took out again, running to 2027 and 2028. The Safe holds 52.7 million OLAS that can move on two signatures, and Olas’s FAQ says Valory’s 10% carries no lock-up.

The developer incentive pool (47.35%) is the largest allocation and the most interesting. It is programmatically distributed to developers whose code is actually used by active agent services. You build a component, register it as an NFT, and earn ETH donations plus OLAS top-ups when your code runs in production. This aligns builders with network growth in a way most protocols don’t.

The LBP raised only $547,580, making this one of the smaller public launches in the space. Only about 1.5% of supply went to public buyers. The recent $13.8 million OTC round (February 2025, led by 1kx) was at the Valory company level.

Olas reports “~77.9M unvested OLAS burnt” in early 2025, and the chain bears it out in substance. One DAO transaction in January 2025 burnedBurnPermanently 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 → 25.2 million OLAS from the treasury and revoked 52.7 million of unvested buOLAS allocations, and those revoked tokens can never leave the buOLAS contract: it would burn them only on a holder’s withdrawal, and with nothing left to release, every such withdrawal reverts. They’re out of circulation permanently, though they still count in OLAS’s total supply. The Mech Marketplace fee burn and the buyback-and-burn from protocol-owned liquidity are deployed but idle: 722.50 xDAI of collected Mech fees sat undrained on Gnosis and the L1 burner held 516.90 OLAS it has never burned.

OLAS is 99.6% below the all-time high of $8.47, set on 3 January 2024. The all-time low was hit in February 2026. Liquidity sits mostly in Balancer pools across Base, Gnosis, Polygon, Arbitrum and Optimism, plus a Uniswap V2 pair, and MEXC is the only centralised venue of consequence. Daily volume is thin, well under $1 million.

How to participate

Run agents via Pearl. Install Pearl (macOS, Linux or Windows), stake OLAS, and launch an autonomous agent. Options include Polystrat (Polymarket trading), Omenstrat (Omen predictions), BabyDegen (DeFi), and Optimus (multi-strategy). Rewards depend on agent performance meeting KPI targets. No GPU required.

Governance. Lock OLAS for veOLAS to vote on staking emissions, protocol 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 →, and treasury allocation. Longer lock periods yield more voting power (maximum four years). Vote via Snapshot (off-chain) or on-chain proposals. No direct yield from governance locking.

Build agents. The most technically demanding participation mode. Use the Open Autonomy Framework (Python, Docker, Tendermint) to build agent services. Register components and agents as NFTs on the on-chain registry. Earn ETH donations and OLAS top-ups when your code is used. Apply to the Olas AcceleratorGPUGraphics 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 → for up to $1 million plus developer rewards. Documentation is solid at stack.olas.network.

Use the marketplace. Hire AI agents for tasks via the Mech Marketplace, or offer your own agents’ services for crypto payments. Currently early-stage with limited turnover.

Honest assessment

What works

The on-chain usage is at least partially verifiable. The transaction and daily-agent counts are figures Olas publishes, and our own read of its registries counts about 5,000 registered services. The 79% prediction accuracy is self-reported. Olas has been on mainnet since July 2023. This isn’t vapourware.

The open-source output is substantial. 146 repositories, 13,698 commits on the primary repo, 15,162 commits on Open AEA. All core code is Apache-2.0 or MIT licensed. Two Code4rena audits completed (December 2023 and January 2026) plus an Immunefi bug bounty. The developer incentive modelModelA trained neural network that takes inputs (text, images, audio) and produces outputs (more text, classifications, generated content). In DeAI the model is the thing that actually does the work.Like a very experienced apprentice who has spent years watching thousands of masters make furniture. They can't explain how they know when a joint is right, but they can make a chair that looks and functions like a Chippendale. The training is invisible. The output is what matters.Model in the glossary →, where protocol-level rewards flow to builders whose code is actually used in production, is structurally sound.

The founding team combines credible academic expertise with practical experience. Minarsch’s Cambridge PhD in game theory applied to multi-agent systems and his years building AEA at Fetch.ai give him domain credibility. Galindo brings cryptography and security research from Birmingham. Pearl demonstrates consumer product ambition, making agent ownership accessible beyond developers.

What doesn’t work

The economics are broken. $110.1K total marketplace turnover by Olas’s count. 99.6% token decline from 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 →. The agent economy generates impressive transaction volume (20.9 million per Olas) but almost no economic value. This is the core problem: Olas has found product-market fit in one vertical (prediction market bots on Gnosis Chain) but hasn’t translated that into sustainable token demand or revenue.

The insider allocation is a structural concern. 42.65% went to founding members and Valory, and the on-chain locks that held most of it all ended by July 2026. The FAQ explicitly states Valory’s allocation has “no locking or vesting requirements.” For a project built on the thesis of community ownership, this contradicts the narrative.

The stack complexity (Python, Docker, Tendermint, multi-agent systems) limits developer adoption compared to simpler alternatives. The Gnosis Chain concentration creates single-chain dependency risk.

The first Code4rena audit found five high-risk vulnerabilities, including a permanent DoS vector and a governance bypass. These have been addressed, but they demonstrate the smart contract complexity and associated risks.

The risk

The token price collapse is the defining risk. At 99.6% below ATH, the incentive model that rewards developers and agent operators in OLAS is severely weakened. If the token stays at these levels, the 47.35% developer incentive pool is worth a fraction of what was originally intended. That isn’t enough to build a thriving developer community.

The prediction market concentration is fragile. If Gnosis Chain loses relevance, or prediction market demand contracts, Olas loses its primary use case and most of its activity metrics. Diversification into DeFi, social, and enterprise agent use cases is essential but unproven.

Low liquidity (a 7-day median turnover of about 1.5% of market cap in September 2026, mostly on DEXDEXDecentralised Exchange. A trading venue where token swaps happen entirely through smart contracts, with no central operator holding user funds. The largest DEXes are Uniswap, Aerodrome, Raydium, PancakeSwap, and Curve.Like a self-service vending machine that lets you swap one type of coin for another. The machine sets the exchange rate based on its current stock, anyone can deposit coins to refill it, and there's no clerk behind the counter.DEX in the glossary → pools) creates execution risk for anyone taking a meaningful position. No major CEX listing beyond MEXC limits retail access.

My position

I don’t hold OLAS. I’m watching the on-chain usage, and the protocol architecture is technically sound. But the 99.6% price decline, the insider allocation now free of any on-chain lock, and the $110.1K total marketplace turnover Olas reports tell a story of a project that has solved the technical problem but not the economic one. I would reconsider if the token stabilises and marketplace turnover reaches meaningful levels. Pearl now runs on Windows as well as macOS and Linux, which removes one of the adoption barriers I was watching.

Freedom Score: 62/100

Olas scores 62/100 (C grade). Full methodology at Freedom Score Methodology.

Infrastructure decentralisation (12/20): Agents run on a permissionless multi-agent stack across nine chains. Agent services use Tendermint for off-chain coordination with on-chain settlement via Safe multisigs. Anyone can register components, deploy agents, and run services without permission. However, Valory operates key infrastructure, the Open Autonomy codebase has a single primary maintainer, and the vast majority of agent activity is concentrated on Gnosis Chain. Cross-chain governance bridges add complexity and potential failure points.

Governance decentralisation (10/20): veOLAS governance runs through on-chain contracts, and the OLAS token is owned by the Timelock, which is correct. Two parties can propose and execute on that Timelock: the current GovernorOLAS (0x060D…51E6), granted its roles on 15 June 2026, and a 5-of-9 Gnosis Safe. Passed votes do execute (proposal 16 did in September), but the Safe can act without a vote. The minimum delay reads zero, so either path acts against the token contract instantly, with no window for holders to react.

A passed vote executed on 13 September 2026 raised the bar for both steps. Proposing now takes 250,000 veOLAS, up from 5,000, and quorum rose from 3% to 10% of veOLAS voting power. Fewer holders can table a proposal on their own, and a low-turnout vote is harder to pass.

The DAO was founded in 2022 with approximately 50 participants, VoteWeighting lets veOLAS holders direct staking emissions, and cross-chain bridges extend voting to six chains. Those are real. On top of them, 42.65% insider tokens (founding members plus Valory), largely claimed into veOLAS voting locks in 2022 and now unlocked, create significant governance concentration, and the participation rate isn’t publicly reported. The score is held here pending the next monthly review, where the on-chain finding gets weighed properly.

Token distribution fairness (7/15): 42.65% insider allocation is a significant concern: about 69% of it was contract-locked at genesis, but every lock had ended by July 2026. Only approximately 1.5% went to the public LBP, raising $547,580. The 47.35% developer incentive pool is the largest allocation but is protocol-controlled, not directly community-held. Taking about 77.9 million OLAS out of circulation in January 2025 (25.2 million burned, 52.7 million revoked and permanently locked) was positive. The LBP mechanism was transparent and fair. No traditional VCVCVenture Capital. Private investors who fund projects at an early stage in exchange for equity or token allocations. VC rounds are typically pre-launch, at steep discounts to any future public price, with multi-year vesting.Like angel investors in a startup who buy shares before the company goes public. They take more risk because the company might fail, so they get a better price. Once the company IPOs they can sell, and the public market pays whatever price it thinks is fair.VC in the glossary → round for the protocol itself.

Censorship resistance (10/15): Agent services are permissionless to deploy and operate. The on-chain registry is open and uncensored. Multi-chain deployment provides redundancy. Safe multisigs provide censorship-resistant agent wallets. However, Valory controls the codebase and could influence agent package availability. Gnosis Chain concentration means most activity depends on one chain’s liveness.

Data sovereignty (10/15): Agent operators maintain sovereignty over their instances and Safe wallets. Pearl enables users to own agents locally. Agent code is open-source and can be self-hosted. Component and agent NFTs are owned by their creators. However, agents depend on external protocol data (Omen, Polymarket), and the Mech Marketplace creates infrastructure dependencies.

Open source and transparency (13/15): All core repositories are open source under Apache-2.0 or MIT. 146 repos, 13,698 commits on the primary repo, 15,162 on Open AEA. Smart contracts verified on Etherscan. Two Code4rena audits completed. Immunefi bug bounty programme. Quarterly blog reports with transparent metrics. Minor deductions for infrastructure code opacity and unpublished 2026 audit results.

Path to improvement

Three changes would materially increase Olas’s score:

1. Re-lock insider tokens. A 42.65% insider allocation with no live lock is the single biggest credibility gap. The 2022 veOLAS and buOLAS locks showed it can be done, but they all ended by July 2026; new on-chain lock-ups on the Safe’s 52.7 million OLAS and the founding members’ released tokens would demonstrate long-term alignment.
2. Diversify beyond Gnosis Chain prediction markets. Olas’s activity sits mostly in Gnosis Chain prediction markets, which is impressive but creates dangerous concentration. Demonstrating agent activity across multiple verticals (DeFi, social, enterprise) and chains would reduce single-point-of-failure risk and broaden the protocol’s value proposition.
3. Grow marketplace economics. $110.1K total marketplace turnover isn’t a functioning economy. The transaction volume (20.9 million per Olas) proves the infrastructure works. The next step is translating that volume into meaningful economic value through higher-value agent services, enterprise use cases, and broader marketplace adoption.

Returns Score: 38/100

OLAS scores 38/100 (F grade). Full methodology at Returns Score Methodology.

Token utility (14/20): Credit where it is due: OLAS has more token utility than most agent protocols. Staking earns rewards tied to actual agent activity via Proof-of-Active-Agent rather than passive lockup. Bonding provides protocol-owned liquidity. veOLAS governance gives holders voting weight on emissions and protocol direction. The utility mechanisms are well-designed on paper. The problem is not the design; it’s the scale at which they operate.

Value accrual (5/20): The marketplace has turned over $110.1K, by Olas’s own counter. Total. Not per day, not per month. Total lifetime turnover. The Mech Marketplace has processed 14.8 million agent-to-agent transactions, but the economic value flowing through those transactions is negligible. The 99.6% token decline from ATH tells you the market’s verdict on the value accrual mechanics. Bonding modelled on Olympus DAO hasn’t created meaningful protocol-owned liquidity at these volumes. The fee-burn and buyback contracts are deployed but have never burned OLAS. The architecture for value capture exists; the value itself doesn’t.

Supply dynamics (7/20): The cap is 1 billion over the first 10 years, then 2% a year. About 529 million had been minted by October 2026, so up to roughly 470 million more can arrive by mid-2032, on average about 15% of today’s supply a year, against demand that hasn’t kept pace. The 42.65% insider allocation is no longer a locked overhang: about 69% of it sat in veOLAS and buOLAS locks from 2022, every one had ended by July 2026, and insiders now hold liquid tokens they can sell at any time. That is dilution rather than a supply shock, which puts the score in the dilutive band; taking about 77.9 million OLAS out of circulation in January 2025 helps at the margin.

Revenue sustainability (5/25): The protocol is almost entirely emission-dependent. Agent rewards come from token emissions, not from external revenue. The marketplace generates trivial fees by its own reported figures. Pearl is free. There is no subscription model, no enterprise licensing, no compute fees flowing back to the protocol. If emissions stopped tomorrow, the economic incentive to run agents would collapse. This is the fundamental challenge: impressive transaction volume built on a subsidy, not on sustainable economics.

Liquidity and access (7/15): Daily trading volume under $1 million, primarily on DEXs. No major CEX listing beyond MEXC. The 99.6% drawdown from ATH suppresses new buyer confidence. Thin liquidity creates severe slippageSlippageThe difference between the expected price of a trade and the price you actually get when the trade executes. Slippage usually goes against the trader and gets worse with bigger trades or thinner markets.Like trying to buy 1000 bananas at the corner shop. The first few are at the marked price, but by the time you've bought them all you've moved the price up because there are no more bananas left at the original level. The shop has to restock at higher cost.Slippage in the glossary → risk; even a modest market sell would move the price meaningfully. For a protocol with substantive technical merit, the market access is remarkably poor.

Path to improvement

Three changes would materially increase Olas’s returns score:

1. Generate real marketplace revenue. The protocol needs transactions that produce fees beyond agent-to-agent messages. Higher-value agent services (enterprise automation, cross-chain DeFi strategies, data analysis) priced in meaningful amounts would create the fee base that makes the tokenomics work. $110.1K of lifetime turnover, by Olas’s count, is a prototype, not an economy.
2. Enforce insider vesting on-chain. Put the insider tokens now free of their 2022 locks, including the 2-of-3 Safe’s holding and Valory’s 10%, under smart contract-enforced vesting with transparent release dates. This single change would eliminate the largest source of structural sell pressure uncertainty and signal credible long-term alignment from the founding team.
3. Secure a Tier-1 CEX listing. Thin DEX liquidity and sub-$1 million daily volume make OLAS practically uninvestable for anyone beyond small retail positions. A Binance or Coinbase listing would dramatically improve price discovery, reduce slippage, and broaden the holder base beyond the current niche audience.

Score change log

Date	Score	Change	Reason
2026-10-07	Returns	35 → 38 (F held)	Supply Dynamics 4 → 7: no insider tokens are locked or unvested any more, so the risk is ~15% a year of emission dilution, the dilutive band rather than supply shock.
2026-10-07	Data	N/A	Insider vesting corrected from chain: about 69% of the 426.5M was contract-locked at genesis (181.0M veOLAS, 111.9M buOLAS), not about a quarter, and every lock had ended by July 2026. Scores unchanged pending review.
2026-10-06	Data	N/A	The 77.9M OLAS cut is 25.2M burned plus 52.7M revoked and permanently locked in buOLAS (still in total supply). The Mech fee burn and buyback have never burned OLAS, so the 2026-06-26 row’s burn attribution was wrong. Scores unchanged.
2026-10-06	Data	N/A	buOLAS described correctly as an enforced yearly release on 113.2M OLAS, ended July 2026. Tokenomics now opens on demand, value capture, live status and dilution. Scores unchanged.
2026-10-04	Data	N/A	Governance updated: a GovernorOLAS vote executed on 13 September 2026 raised the proposal threshold from 5,000 to 250,000 veOLAS and quorum from 3% to 10%, read on chain. Scores unchanged.
2026-09-29	Data	N/A	Liquidity finding corrected: a Uniswap V2 pool on Ethereum now carries about half of daily volume, with MEXC the largest centralised venue at about a quarter. Scores unchanged.
2026-09-24	Data	N/A	Supply restated to the on-chain 528.8M (52.9% of cap) with CoinGecko circulating now bound, two leftover $89K turnover figures bound to Olas’s counter, liquidity venues refreshed, and Olas’s counters no longer called verifiable. Scores unchanged.
2026-09-24	Data	N/A	Governance corrected: our August trace checked two retired GovernorOLAS deployments and missed the current one, which has held Timelock proposer and executor roles since 15 June 2026 alongside the 5-of-9 Safe. Governance 10/20 held, flagged for the monthly.
2026-09-24	Data	N/A	Olas’s published counters had moved (20.6M transactions, 351 daily agents, $109.6K turnover on 24 Sep); now bound to a weekly refresher. The unpublished Gnosis Safe share is dropped. Scores unchanged.
2026-09-24	Data	N/A	Robinhood Chain added: DAO proposal 16 executed on Ethereum, adopting the Olas instance there as canonical, and a read of the chain shows 17 marketplace requests so far. Scores unchanged.
2026-09-23	Data	N/A	Integrity records re-cut: 2 claims now quote Olas’s own pages verbatim and 7 are our own measurements. Scores unchanged.
2026-09-07	Data	N/A	WS3b censorship rail: OLAS is owned by a TimelockController whose getMinDelay returns zero, with a 5-of-9 Safe as proposer and executor. Governance can act with no notice window, which our “bridges could be paused by governance” line does not convey. Censorship Resistance 10/15 held, anchored.
2026-08-15	Data	N/A	Circulating supply de-hardcoded and bound to the live source. The stated 47.2% was 14.4pp stale against 61.61% of total supply.
2026-08-13	Data	N/A	Olas’ live counters have drifted from our recorded figures: 598 daily active agents (we say >600), 14.23M A2A transactions (we say 14.5M), and the Safe share is published as a Gnosis co-founder quote saying “over 75% of Safe transactions on Gnosis Chain” on many days, not 35%. Refresh at the monthly.
2026-08-13	Data	N/A	Traced the OLAS governance permissions on-chain: the Timelock owns the token, but its only proposer and executor is a 5-of-9 Safe and both GovernorOLAS deployments have had their roles revoked, with a zero minimum delay. Corrected the Governance section; the score is flagged for the monthly review.
2026-08-13	Data	N/A	Confirming positives from the same pass: the 1B ten-year cap is enforced in the contract (supply 528.79M plus inflationRemainder 471.21M sums exactly to the cap) and every contract in the ownership path is source-verified. Migrated Olas onto the verdict spine.
2026-06-26	Data	N/A	DefiLlama LlamaAI harvest cross-check, verified independently (api.llama.fi + CoinGecko): circulating supply fell to 223.3M (47.2% of 473.6M total) from the 237.0M/50.05% cited at the June review, consistent with the Mech-fee and buyback burns. ATH decline holds at -99.6% (verified -99.64% from $8.47, 3 Jan 2024). Refreshed stale daily volume ($228K → ~$133K). Score unchanged (marginally less circulating supply does not move Supply Dynamics).
2026-06-02	Data	N/A	Supply figures aligned to CoinGecko (June 2026): total 527.9M → 473.5M, circulating 236.2M → 237.0M, percentage 44.75% → 50.05% (the shift is the total-supply denominator; circulating tokens barely moved). TokenFactsStrip + body updated. Score unchanged. June monthly review.
2026-05-03	Editorial	N/A	Pearl OS support corrected (now macOS, Linux and Windows; was previously listed as macOS-only). Daily active agents refreshed 542 to 600+ per OLAS dashboard. No score change.
2025-03-02	Both	N/A	Initial publish. Freedom 62/100, Returns 35/100.