Golem Network

What it does

Golem Network is a peer-to-peer marketplace for computing power, one of the oldest decentralised compute projects in existence. Conceived in 2014, presented at DEVCON0, and funded via a legendary 2016 ICOICOInitial Coin Offering. A token sale where a project sells tokens directly to the public, usually before any product exists. ICOs dominated 2017-2018 funding and are now mostly replaced by airdrops, IDOs, or fair launches.Like a company selling shares to the public before going public, except with no SEC oversight, no audited financials, and often no product at all. The 2017 ICO boom showed why those guardrails exist in traditional finance.Read more → that raised 820,000 ETH (~$8.6M) in under an hour, Golem connects Providers (who share spare CPU/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.Read more → resources) with Requestors (who consume compute) using the GLM 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 → as the payment medium.

The core software is Yagna, an open-source Rust framework that handles task negotiation, execution, and payment. Providers run Yagna on their own hardware. There are no centralised data centres. Requestors submit jobs which are split into subtasks and distributed across provider nodes for parallel processing.

Supported workloads include 3D rendering/CGI (the original use case), AI 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 → 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.Read more →, scientific simulations, zero-knowledge proofZKZero Knowledge. A class of cryptographic proofs that let you prove something is true without revealing any of the underlying information. ZK lets a network verify a transaction without seeing the transaction's contents.Like proving you know the password to a safe by demonstrating you can open it, without ever saying the password out loud. The verifier learns that you know the password and nothing more.Read more → generation (via Satori integration), and general-purpose batch compute. Pricing runs approximately $0.001/core-hour for CPU and $0.01/core-hour for GPU, claimed to be 70-90% below AWS equivalent.

Payments operate on Ethereum mainnet and Polygon 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 → (added 2021). Yagna v0.16.0 (August 2024) introduced a 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 → deposit mechanism allowing end-users to pre-pay for compute through a lock contract.

Founded by Julian Zawistowski, Piotr Janiuk, Aleksandra Skrzypczak, and Andrzej Regulski. The entity is Golem Factory GmbH, registered in Zug, Switzerland. In 2019, Zawistowski and Regulski departed to establish the Golem Foundation, a separate non-profit entity pursuing “riskier” research. Over $40M in ETH and GNT was transferred from Factory to Foundation, a significant leadership and treasury split.

Current leadership: Piotr Janiuk (CEO, co-founder, mathematics/CS, University of Warsaw), Przemyslaw Rekucki (CTO), Pawel Burgchardt (CPO), Arkadiusz Cybulski (Head of AI/GPU).

Brass Golem (2018) was one of the first Ethereum mainnet applications. The token migrated from GNT to GLM (ERC-20, 1:1 swap, no deadline) in 2020.

Value proposition

The OG decentralised compute network with ten years of continuous development. Golem’s value proposition is simple: permissionless, censorship-resistant computing power at 70-90% below cloud provider pricing.

Recent expansion: Salad partnership, Arkiv, Ecosystem Fund

The most commercially significant recent development is the Salad.com partnership (January 2026). Salad is a GPU cloud platform that claims $200M+ in annual revenue, and the engineering test mirrors Salad’s commercial workloads (AI inference, 3D rendering, drug discovery simulations) onto Golem’s permissionless execution layer. If this progresses beyond testing, it represents Golem’s first real commercial integration at scale.

Arkiv extends Golem’s reach: a decentralised database and data availability layer operating as an L2 on Ethereum with L3 DB-Chains, using GLM for data writes. This is a meaningful extension of GLM utility beyond compute.

The Ecosystem Fund has supported six projects including Web3 Pi (Ethereum nodes on Raspberry Pi), Satori Network (ZK proving, mainnet integration July 2025), and Keccak Optimization (ZK proof research).

The counter-narrative: zero revenue, slowing development, leadership split

The counter-narrative is severe: Golem has zero protocol revenue. The protocol charges 0% commission on provider earnings. GLM earned by providers is sold immediately, creating high token velocity with no value capture. After ten years of building, the protocol has no sustainable revenue model. The team’s financial sustainability depends on treasury assets and the Foundation’s holdings, neither of which is transparently disclosed.

Development velocity has slowed. The yagna core repo shows no commits since October 2025. Modelserve AI (the key AI product) has been temporarily suspended. Blog frequency dropped to roughly one post per quarter. Meanwhile, Akash, Render, and io.net have all gained more market traction in the AI compute narrative.

The 2019 leadership split is concerning. Two entities (Factory and Foundation) with shared token but divergent strategies and separate treasuries. The Factory builds the protocol; the Foundation pursues research. Coordination between them is unclear.

For the sovereignty thesis, Golem scores well. The network is explicitly designed as censorship-resistant computing. Anyone can join without permission or KYC. All core code is GPL-3.0 licensed with 49 contributors on the yagna repo. The 82% ICO distribution is one of the fairest in crypto history. The architecture is verifiably permissionless and decentralised.

Tokenomics

GLM has a fixed supply of 1 billion tokens, 100% circulating. No 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.Read more →, no future 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.Read more →, no remaining unlocks, no 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 →. 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 → equals market cap. This is about as clean as token supply gets.

The 2016 ICO distributed 82% of tokens to participants, extraordinary by any standard. Golem Factory received 12%, and the team received 6%. All tokens were immediately liquid.

GLM is used for compute payment (sole currency between requestors and providers), smart contract deposits for pre-payment (v0.16), and Arkiv data writes. There’s no 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 → mechanism, 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, no governance token weighting, and no fee-sharing. The protocol charges 0% commission.

The token velocity problem is structural. Providers earn GLM for compute and sell it for operating costs. Without staking lock-ups, burn mechanics, or protocol fee capture, there are no supply-side sinks. Token value depends entirely on compute demand exceeding sell pressure from providers.

GLM is 84% below its 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 → of $0.86 (November 2021). Listed on 44 exchanges including Binance, Coinbase, OKX, KuCoin, and Upbit. 19,813 holders on Ethereum, remarkably low for a 2016-vintage token.

How to participate

Become a Provider. Share spare CPU/GPU resources. Install Yagna, configure hardware, earn GLM from requestors. Permissionless, no KYC. Technical skill: intermediate.

Submit Compute Jobs. Use the Yagna SDKSDKSoftware Development Kit. A collection of code libraries, documentation, and tools that lets developers integrate a service into their applications without writing everything from scratch. SDKs are how projects become easy to build with.Like a plug-and-play kit for building furniture. You don't have to mill your own wood, forge your own screws, or design the joinery from scratch. The kit gives you pre-cut parts and instructions so you can assemble the thing in an afternoon.Read more → (Python, TypeScript) or CLI to distribute tasks across the network. Pay in GLM. Pricing ~$0.001/core-hour CPU, ~$0.01/core-hour GPU. Technical skill: intermediate to advanced.

Build on Arkiv. Use the L2/L3 data availability layer for decentralised database applications. GLM used for data writes. Technical skill: advanced.

Honest assessment

Freedom Score: 69/100

Ten years in, Golem is still one of the most honestly built compute networks in crypto. The token distribution and open-source credentials are hard to argue with.

Infrastructure Decentralisation: 14/20. Permissionless P2P network: anyone can join as provider or requestor without permission. Providers run their own hardware. Payments on Ethereum and Polygon. Relay infrastructure and task matching involve some centralised coordination points. Provider count is modest compared to competitors like Akash or io.net.

Governance Decentralisation: 10/20. GAPs (Golem Amendment Proposals) exist as a governance mechanism since 2020. In practice, Golem Factory GmbH retains control over protocol development, roadmap, and treasury. The 2019 split created two entities which distributes power somewhat, but neither is community-governed. No formal 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 →. Ecosystem Fund allocation is team-directed.

Token Distribution Fairness: 11/15. 82% of tokens went to ICO participants, one of the fairest distributions in crypto history for a 2016 project. Only 18% to team/company. Fully circulated with no remaining unlocks. The $40M+ transfer to Foundation concentrates significant resources, but the initial distribution was exemplary.

Censorship Resistance: 11/15. Explicitly designed as censorship-resistant computing. Permissionless participation without KYC. Supports arbitrary workloads. However, providers choose which tasks to accept, and the network’s small scale means a coordinated attack is feasible. No specific anti-censorship technology like encrypted task delegation.

Data Sovereignty: 10/15. Requestors send tasks to providers, so data leaves the requestor’s machine. Plans for VPN capabilities and SGXSGXIntel Software Guard Extensions. The first widely-deployed TEE technology, introduced in 2015. SGX creates encrypted memory regions (enclaves) where code and data are protected from the operating system and the machine's owner.Like a safe deposit box at a bank. The bank owns the safe room and can see who comes in and out, but they can't see what's inside the boxes. SGX gives applications a private box on a shared computer.Read more → exist but are not in production. No TEETEETrusted Execution Environment. A hardware-secured region of a CPU or GPU where code runs in isolation, so even the machine's operator can't read what's happening inside. TEEs give decentralised AI inference privacy guarantees.Like a bank vault inside a bank. The bank owns the building, staffs the lobby, and runs the security cameras. But what's inside the vault is invisible to everyone, including the bank staff, unless the customer opens it.Read more → or confidential computingConfidential ComputeHardware-enforced computation where data and code are encrypted in memory and only the authorised application can access them. The machine's operator cannot read what the application is doing even though they own the machine.Like renting space in a bank vault. The bank owns the building and runs the security, but what you put in the vault is invisible even to the bank staff. Only you have the key.Read more → currently deployed. Arkiv extends data sovereignty concepts. Current data privacy during compute is limited.

Open Source Transparency: 13/15. All core code GPL-3.0 licensed. 100+ repos, 49 contributors on yagna. OpenZeppelin audited the token contract with a strong rating. Team is identifiable and public-facing. Development cadence has slowed (5-month gap on yagna repo). Foundation activities are less transparent than Factory’s.

Returns Score: 46/100

Token Utility: 10/20. GLM functions solely as a payment medium: requestors pay providers in GLM. The deposit mechanism adds smart contract locking for pre-payment. Arkiv extends utility to data writes. However, there’s no staking requirement, no governance voting weight, no fee-sharing mechanism. The token is a pure medium of exchange, relatively thin utility compared to tokens with staking, governance, and burn mechanics.

Value Accrual: 6/20. GLM has no burn mechanism, no staking rewards, no protocol fee capture, and no revenue sharing. The token’s value depends entirely on demand for compute priced in GLM. With no supply-side sinks, token velocity is high: providers earn and sell immediately. The deposit mechanism creates temporary lock-up but not permanent value accrual. FDV = market cap (no dilution), which is the sole positive.

Supply Dynamics: 14/20. Strong fundamentals: fixed 1B supply, no inflation, no future emissions, no remaining unlocks. FDV = market cap. 82% distributed to public. All vesting complete (there never was any). Clean structure. Deducted for: no deflationary mechanism, no staking lock-up, and ~20% held by team/foundation entities.

Revenue Sustainability: 5/25. Golem charges 0% commission on provider earnings. The protocol captures zero revenue from usage. Pricing at $0.001-$0.01/core-hour is exceptionally low. The Ecosystem Fund finances development but its size and runway are not disclosed. The Foundation holds significant assets but this is a separate entity. Salad partnership is an engineering test, not a revenue contract. Without protocol-level revenue capture, sustainability depends on treasury management.

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: 11/15. Listed on 44 exchanges including Binance, Coinbase, OKX, KuCoin, and Upbit. Excellent coverage. Daily volume provides adequate liquidity. Available on Uniswap for decentralised access. Polygon bridgeBridgeA protocol that lets you move assets from one blockchain to another. Bridges typically lock the asset on the source chain and mint a wrapped version on the destination chain. Bridges are notoriously the most-attacked component in crypto.Like a coat check at a club. You hand over your coat, get a numbered ticket, and the club promises to return the coat when you bring back the ticket. The trust assumption is that the coat check doesn't lose your coat or run away with it.Read more → available.

Quadrant: B (High Freedom, Low Returns)

Golem sits firmly in Quadrant B: one of the strongest sovereignty plays in decentralised compute with demonstrable permissionless infrastructure and an exemplary token distribution, but fundamentally weak value capture that limits investment returns.

Key risks

• Zero protocol revenue. 0% commission means the protocol captures no value from ten years of network usage. Structurally unsustainable without treasury subsidies.
• Token velocity problem. No staking, no burn, no lock-up incentives. GLM earned by providers is sold immediately, creating persistent sell pressure.
• Development stagnation. Yagna core repo shows no commits since October 2025. Modelserve AI suspended. Blog frequency dropped to quarterly.
• Leadership fragmentation. The 2019 split divided resources, talent, and $40M+ in treasury between Factory and Foundation. Coordination unclear.
• Competitive erosion. Akash, Render, and io.net have all gained more traction in the AI compute narrative despite Golem’s 10-year head start.
• Low holder count. 19,813 holders for a 2016-vintage token suggests limited and declining retail interest.
• Salad dependency. The most promising commercial development is currently only an engineering test, not a committed production deployment.
• 84% from ATH. Token peaked at $0.86 in November 2021 and has been in steady decline.

The Salad.com partnership is the most concrete commercial signal in years. If it converts from engineering test to production deployment, the narrative changes. Until then, Golem is a project that built the right thing at the right time and hasn’t worked out how to get paid for it.

My position. I hold GLM, staked to a router operator. The sovereignty thesis is the draw: permissionless infrastructure, GPL-3.0 licensed code, one of the fairest token distributions in crypto history. The returns case is weak by design, because Golem charges zero commission. I’m not holding it for yield. I’m holding it because the Salad.com partnership is the first real commercial signal in years and I want exposure if it converts to production. Zero protocol revenue remains the defining problem.

Score change log