FLock.io

What it does

FLock.io is a decentralised AIDeAIDecentralised 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.Read more → model 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.Read more → and validation platform that combines federated learning with blockchain. The core promise: train AI models collaboratively without sharing raw data. Only 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.Read more → updates (gradients) are transmitted; the training data never leaves the participant’s device.

The architecture has three layers:

• AI Arena (train.flock.io): the live training platform on Base. Training Nodes submit models, Validators evaluate them via comparative scoring, and Delegators stake tokens to nodes and validators. On-chain smart contracts handle rewards and slashing.
• FL Alliance: the federated learning layer where participants fine-tune consensus models using their private local datasets. Still in testnet as of mid-2025. This is the privacy-preserving component that makes FLock genuinely different.
• Moonbase / APIAPIApplication Programming Interface. A structured way for one piece of software to talk to another. In DeAI, APIs let applications request inference from a model without running the model themselves.Like a waiter in a restaurant. You don't walk into the kitchen and cook your own meal. You tell the waiter what you want, they tell the kitchen, the kitchen cooks it, and the waiter brings it back. The API is the waiter.Read more → Platform: deployment layer launched September 2025. An OpenAI SDK-compatible API for deploying and monetising trained models, with revenue sharing for contributors. Moonbase also introduced FOMO (FLock Open Model Offering) as part of Tokenomics v3: fair-launch model tokens backed by hosting services on FLock API. Model tokenholders get discounted usage or can stake for yield.

The technical innovation is FLoRA: using Low-Rank Adaptation (LoRA) for efficient fine-tuningFine-tuningThe process of taking a pre-trained model and training it further on a smaller, more specialised dataset to adapt it to a specific task, domain, or style. Fine-tuning is much cheaper than training from scratch.Like hiring an experienced general practitioner doctor and giving them six months of focused training in a sub-speciality. You don't have to teach them medicine from scratch. You just narrow their expertise to the area you actually need.Read more → with minimal compute overhead, replacing central aggregator servers with blockchain-based peer review. FLock achieved the first empirical validation of fine-tuning a 70B LLMLLMLarge Language Model. A neural network trained on vast amounts of text to predict the next word in a sequence. Modern LLMs (GPT, Claude, Llama, Qwen, DeepSeek) generate human-quality text and are the foundation of most modern AI products.Like an autocomplete that read every book ever written. It has no memory of individual texts but it has absorbed the patterns of language so deeply that it can generate paragraphs that sound human. The skill is statistical, not conscious.Read more → in a decentralised multi-domain setting.

Founded by Jiahao Sun (MSc Computer Science, Oxford; Honorary Research Fellow, Imperial College London; former Director of AI at RBC Wealth Management). Zehua Cheng (Chief AI Scientist, DPhil candidate at Oxford) and Dr. Zhipeng Wang (Lead Blockchain Researcher, leading Ethereum Foundation-funded research) round out the technical leadership.

The entity is FLOCK.IO LTD, registered in London (Companies House #14039622, incorporated April 2022). Total funding: ~$9M, split between seed ($6M led by Lightspeed Faction and Tagus Capital, with DCG, OKX Ventures, Volt Capital) and strategic ($3M from Animoca Brands).

The academic credentials are strong: NeurIPS 2022 Best Paper Runner-up, NeurIPS 2023 Best Technical Demo, IEEE GBC 2025 Best Application Award, sole AI infrastructure recipient of an Ethereum Foundation Academic Grant (2024), and CB Insights AI 100 (2025).

Value proposition

FLock addresses a real problem. Most AI training requires centralising data, a privacy, sovereignty, and regulatory nightmare. Federated learning trains models where the data lives. FLock adds blockchain incentives (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.Read more →, slashing, rewards) to make this economically viable and verifiable.

Usage signals: revenue, participation, but tiny task volume

The usage metrics from the first 10 months tell a mixed story:

• Protocol fee revenue: 11.9M FLOCK (~$2.7M). On-chain revenue from Base smart contracts, verifiable on-chain.
• 784,137 validation submissions vs 9,062 training submissions, showing meaningful validatorValidatorA computer that runs the full blockchain protocol, verifies transactions, and proposes new blocks. Validators are the workers that keep a Proof of Stake network running, and they earn rewards for doing the work correctly.Like a notary public who witnesses and stamps legal documents. Validators witness transactions, check they follow the rules, and stamp them into the permanent record. A notary who commits fraud loses their license. Validators work the same way, except the license is staked tokens that get slashed on misbehaviour.Read more → activity.
• 196 training nodes including 6 top-50 Kagglers, a quality signal.
• 262 validators and 1,416 delegators, showing substantive participation.
• 55,928 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.Read more → holders.

But: only 16 training tasks completed in that period. Per-task economics are impressive ($169K revenue per task) but the volume is tiny. Task creation is still centralised: the FLock team decides what gets trained. Community task creation is planned but not shipped.

The gmFLOCK staking mechanism (Tokenomics v2, launched April 2025) is well-designed. Stake FLOCK to receive non-transferable gmFLOCK, required for AI Arena participation. Lock up to 365 days; longer lock means more gmFLOCK. 25% of 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.Read more → is voluntarily locked at an average of 265 days. That’s demonstrable long-term alignment.

The healthcare use case, including a partnership with Moorfields Eye Hospital for ophthalmology diagnostics and cross-continental collaboration on blood glucose 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.Read more → for diabetic patients, demonstrates the real-world applicability of the approach.

The counter-narrative: testnet status, closed-source, locked supply

The counter-narrative: FL Alliance, the core privacy-preserving federated learning layer, is still in testnet. The API Platform launched only in September 2025. Core platform code is partially closed-source. No 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.Read more → audit has been published. And 68.5% of token supply is still locked, with significant future dilution ahead.

Partnerships with Qwen (Alibaba Cloud) and Animoca Brands are at the MoU stage, not revenue-generating agreements. More substantive is FLock’s role as “AI Strategic Partner” in the UNDP SDG Blockchain 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.Read more → (second cohort, announced August 2025 alongside Stellar), mentoring five pilot projects across climate finance, energy access, healthcare, and supply chain. FLock participated in the UNDP High-Level Strategic Dialogue at UN Headquarters. That’s a credibility signal that carries weight beyond crypto-native partnerships.

FLock also partnered with Walrus Protocol and Sui Foundation. Walrus provides decentralised encrypted storage for gradients, and FLock is working on fine-tuning an open-source model for the Sui ecosystem. This expands the chain footprint beyond Base.

Tokenomics

FLOCK has a fixed supply of 1 billion tokens with approximately 315 million circulating (31.5%). 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.Read more → was December 31, 2024 on Base.

Distribution: community incentives 47%, team & shareholders 22.5%, treasury/ecosystem 18%, community 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.Read more → 12.5%. The 66.7% community vs 33.3% insider split is better than most. Team and investor tokens have a 1-year 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.Read more → + 2-year linear 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.Read more →. Community tokens mint over 60 months with 1% monthly decay.

The gmFLOCK mechanism creates actual lock-up: stake FLOCK to receive non-transferable gmFLOCK (required for participation). Base lock of 30 days (1:1 ratio), scaling up with longer commitments (+0.006 per additional day). 25% of circulating supply is voluntarily locked at an average of 265 days.

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.Read more → mechanism exists. Value accrual comes through protocol fee revenue ($2.7M in 10 months) and planned revenue sharing via the API Platform. How fees flow to token holders vs the protocol treasury isn’t explicitly documented.

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.Read more →/MCap ratio of 3.16x signals meaningful future dilution; 31.5% of supply is circulating. Listed on Gate.io, Bybit, KuCoin, MEXC, Bitget, and CoinEx. 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.Read more → was in September 2025. Next unlock: 31 March 2026, releasing 16.34M FLOCK (1.6% of supply).

How to participate

Run a Training Node. Submit AI models to training tasks. Stake FLOCK via gmFLOCK. Includes top-50 Kagglers among current participants. MLMLMachine Learning. The branch of AI where systems learn patterns from data instead of being explicitly programmed with rules. Modern AI (LLMs, image generation, recommendation systems) is almost entirely machine learning.Like teaching a child to recognise dogs by showing them thousands of pictures of dogs, instead of writing down a precise rulebook for what makes a dog. The child learns the pattern from examples rather than from instructions.Read more → expertise required. Technical skill: advanced.

Stake as Validator or Delegator. Validators evaluate training submissions. Delegators stake to trusted nodes and validators. Lock FLOCK via gmFLOCK for up to 365 days. Technical skill: intermediate.

Deploy Models. Use the API Platform (Moonbase) to deploy trained models. Revenue sharing for contributors. Elite Trainer Programme offers 50% revenue share. Technical skill: intermediate.

Honest assessment

Freedom Score: 55/100

FLock has a federated learning architecture inherently designed for data privacy, strong academic credentials, and decent community token distribution. But task creation is centralised, core code is partially closed, and governance is aspirational.

Infrastructure Decentralisation: 11/20. 196 training nodes and 262 validators, providing verifiable multi-party training. Task creation remains centralised to the FLock team. FL Alliance (the privacy-preserving federated learning layer) still in testnet. Smart contracts on Base inherit Ethereum security.

Governance Decentralisation: 9/20. 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.Read more → governance described in docs with weighted/quadratic voting. In practice, team controls task creation, protocol upgrades, and treasury. A decentralisation plan was due Q2 2025, but no evidence it was published. Governance is more described than demonstrated.

Token Distribution Fairness: 10/15. 66.7% community vs 33.3% team/investors, a good ratio. 1-year cliff + 2-year linear vest for team. Community tokens minted over 60 months with decay. 5% airdrop to testnet participants. ~31.5% circulating after 14 months is a moderate unlock pace.

Censorship Resistance: 8/15. Federated learning means data never leaves participants’ devices, a strong privacy architecture by design. Base L2L2Layer 2. A blockchain that runs on top of an L1 to provide cheaper or faster transactions while inheriting the L1's security. L2s batch many transactions and post compressed proofs back to the L1.Like an express lane built on top of a busy motorway. The express lane handles its own traffic at high speed, but it still feeds back into the main motorway and uses the motorway's bridges and tolls for security.Read more → contracts are immutable once deployed. However, task creation is centralised and the API Platform is team-operated, creating single points of control for what gets trained and served.

Data Sovereignty: 10/15. Core value proposition is data staying with owners. FL Alliance designed for privacy-preserving training without data movement. Healthcare demo validated cross-continental collaboration without data sharing. FL Alliance still in testnet, so the full vision isn’t yet delivered.

Open Source Transparency: 7/15. 34 GitHub repos with some actively maintained. Training node quickstart and validator code are open source. Main FLock repo is near-empty (README + LICENSE only). Core platform code and smart contracts not fully public. Academic papers peer-reviewed at NeurIPS and IEEE add credibility.

Returns Score: 58/100

Token Utility: 14/20. FLOCK required for staking as trainer, validator, or delegator. gmFLOCK mechanism adds meaningful lock-up incentive (up to 365 days, 25% of supply voluntarily locked). Governance voting requires staking. Slashing for malicious behaviour creates concrete skin-in-the-game. Task creation will require tokens (planned). No burn mechanism.

Value Accrual: 13/20. Protocol generated 11.9M FLOCK ($2.7M) in fee revenue in the first 10 months, on-chain revenue from Base smart contracts, not inflationary rewards. API Platform has revenue sharing for model contributors. Elite Trainer Programme offers 50% revenue share. Unclear how fees flow to token holders vs protocol treasury. No explicit buybackBuybackUsing protocol revenue to purchase tokens on the open market, usually to burn them or return them to a treasury. Buybacks convert business income into upward pressure on the token by reducing circulating supply.Like a public company using profits to repurchase and retire its own shares. The cash leaves the company's balance sheet, the share count drops, and every remaining shareholder owns a slightly bigger slice of the same business.Read more → or burn.

Supply Dynamics: 14/20. 1B fixed supply with no additional inflationEmissionsNew 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.Read more →. Community tokens minted over 60 months with 1% monthly decay, a controlled emission schedule. 25% of circulating voluntarily locked via gmFLOCK (average 265 days). Team/investor cliff-locked for 1 year + 2-year linear vest. However, 68.5% of supply still locked, creating significant future dilution pressure.

Revenue Sustainability: 9/25. $2.7M protocol fees in 10 months is promising, but from only 16 training tasks, a small volume. Task creation still centralised, limiting organic demand. UNDP partnership is active (AI Strategic Partner in SDG Blockchain Accelerator); Animoca and Qwen partnerships at MoU stage. API Platform launched September 2025, too early for revenue assessment. Revenue depends heavily on team-created tasks; community task creation not yet live.

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.Read more → & Access: 8/15. Listed on Gate.io, Bybit, KuCoin, MEXC, CoinEx, Bitget, providing decent CEX coverage for a small-cap project. 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.Read more → liquidity on Uniswap Base and Aerodrome. Volume adequate but not deep. No Binance or Coinbase listing. Base chain primary, with Walrus/Sui integration expanding the chain footprint.

Quadrant: B (High Freedom, Low Returns)

FLock sits in Quadrant B: genuine data sovereignty innovation with early commercial traction, but too small and too early to demonstrate sustainable returns.

Key risks

• 16 training tasks in 10 months. Revenue per task is strong ($169K) but the volume is tiny. Can it scale 100x?
• Centralised task creation. The team decides what gets trained. Community task creation is planned but not shipped.
• 68.5% of supply locked. Significant future dilution as team, investor, and community tokens unlock over the next 3+ years.
• FL Alliance still in testnet. The core privacy-preserving federated learning layer (the key differentiator) isn’t yet in production.
• Core platform partially closed-source. Smart contracts and platform logic not fully public despite open source claims.
• No smart contract audit. A protocol handling staking and slashing with no published security audit is a concern.
• Small market cap. Vulnerable to large sells. Volume means limited exit liquidity for meaningful positions.
• Competition. Bittensor, Gensyn, and others compete for decentralised AI training. FLock’s federated learning approach is differentiated but unproven at scale.
• Partnership execution. UNDP is active, but Qwen and Animoca partnerships remain at MoU stage, not revenue-generating.

The core question for FLock is whether 16 training tasks in 10 months becomes 160 in the next 10. If FL Alliance ships to mainnet and community task creation launches, the per-task economics suggest real upside. If not, the $2.7M in protocol revenue is a one-off data point from a still-centralised system.

Score change log