AI+Crypto: New赛道 and New Opportunities
TechFlow Selected TechFlow Selected
AI+Crypto: New赛道 and New Opportunities
The integration of AI and cryptocurrency technologies brings transformative opportunities to various industries.
Navigating Web3 tides with focused insightsBy: Trustless Labs
Summary
AI applications integrated with blockchain technology are gaining increasing market attention. ChatGPT acquired hundreds of millions of users in a very short time, and NVIDIA's stock price rose eightfold in 2023, surpassing a market capitalization of one trillion dollars. AI has become the dominant mainstream trend for humanity, and its narrative and capital are now spilling over into related fields, particularly those combining blockchain technology.
Crypto, as a crucial complement to AI, offers broad integration potential and significant development prospects, currently at an early stage. Many institutions and practitioners remain cautious about combining AI with Web3, primarily exploring crypto’s most direct use case—“assetization”: assetizing computing power (cloud computing, compute markets), models (AI Agents), and data (storage).
Currently, while the decentralized nature unique to crypto technology does not improve efficiency or reduce costs during AI training—and still requires substantial infrastructure development—assetization reduces market transaction friction and unlocks previously underutilized computing resources, proving profitable amid widespread compute scarcity. Model assetization allows communities to own or use AI in a decentralized manner; this sector is relatively simple to implement with low barriers to entry, aligning the democratization-of-AI narrative with anxieties around centralized AI control. Decentralized data can moderately reduce costs, but even within crypto contexts, data assetization faces major hurdles and demands extensive business model exploration.
Although there is no full market consensus yet on AI + Crypto, the landscape is already clearly defined. Trustless Labs has analyzed over 60 projects across primary and secondary markets. This article provides a horizontal analysis across seven modules—compute cloud, compute markets, model assetization and training, AI Agents, data assetization, ZKML, and AI applications—to explore the future of AI in the cryptocurrency space and uncover investment opportunities.
Compute Cloud
With the overall growth of the AI market represented by companies like NVIDIA and the ongoing shortage of GPU supply, compute clouds and related compute market projects have become among the first to benefit from AI industry expansion. Compute clouds aim to solve GPU shortages by aggregating idle computing resources from existing cloud providers into a centralized network, reducing compute costs compared to traditional cloud services. These projects vary in focus, with most concentrating on GPU compute, though differing in technical approaches—some based on container technologies like Akash Network, others offering virtual machine-level implementations.
For compute cloud projects, user and business growth form the foundation of value appreciation. These projects are typically viewed as high-PE AI compute enterprises since they do not rely heavily on decentralization narratives. Given that GPU resource scarcity may persist long-term, assessing the commercial capabilities of early-stage participants becomes especially important.
Akash Network
Akash Network is a decentralized cloud computing marketplace built on Cosmos SDK and Tendermint, aiming to utilize underused resources in data centers. Founded in 2015, Akash Network became one of the earliest projects in this field, initially supporting only CPU computing resources. In September 2020, Akash launched its mainnet supporting CPU computing, followed by a successful mainnet upgrade in September 2023 adding GPU support, significantly expanding its service scope and performance. Akash was designed to capture the vast market of underutilized data center resources.
Akash Network employs a unique reverse auction mechanism, allowing users to select the lowest-bidding provider, reducing cloud computing costs—its average prices are approximately one-third of mainstream providers such as AWS and Azure. This pricing strategy not only saves users significant expenses but also enhances market competitiveness and appeal.
The project’s underlying architecture uses the Akash Container Platform built on Kubernetes, orchestrating and managing containers to meet diverse cloud computing needs. This design enables flexible deployment and management, improving resource utilization and flexibility.
In 2020, Akash Network raised $2 million in funding led by Cypher Capital, signaling market recognition of its technology and business model. Founder Greg Osuri brings extensive entrepreneurial experience and cloud architecture expertise since 2008, with Adam Bozanich serving as Co-founder & CTO.
Currently, $AKT has a market cap of $130 million and a fully diluted valuation (FDV) of $220 million.
Ritual
Ritual is a cutting-edge project integrating artificial intelligence (AI) with smart contract technology, building an open AI infrastructure network. The project is driven by an experienced team including Eva Zhang, who previously worked at Socket, Apple, Google, and conducted machine learning research at Stanford University; along with co-founders Niraj Pant and Akilesh Potti, both with backgrounds at Polychain. Ritual aims to enable developers to seamlessly integrate AI into any on-chain application or protocol, including fine-tuning, monetization, and inference execution, promoting fully transparent DeFi, self-improving blockchains, autonomous agents, and generative content.
Though still in early stages, Ritual has achieved several milestones, including launching its first product phase, Infernet. Infernet is a lightweight library enabling computation to be brought on-chain, compatible with any EVM-compatible chain, allowing smart contracts to natively access various AI models for on-chain use cases and tasks. Beyond optimizing decentralized oracle networks, Infernet’s SDK enables DApps to implement features like DAO proposal risk detection and NFT image generation. Ritual plans to launch its second-phase product—Ritual Chain—in the coming months, a sovereign chain with a custom virtual machine dedicated to advanced AI-native applications.
Despite lacking a testnet, economic model, token issuance, or staking mechanism, Ritual’s open-source code, team background, technical roadmap, strategic direction, and community-driven ethos demonstrate strong potential and industry interest. Additionally, Ritual successfully raised $25 million in funding led by Archetype, with participation from Accomplice and Robot Ventures, reflecting investor confidence in its vision and technology.
Render Network
Render Network is a decentralized GPU rendering platform leveraging globally distributed high-performance GPU nodes to provide scalable, on-demand rendering capacity for artists and developers. While focused on rendering, it also supports AI computing, advancing next-generation rendering and AI technologies. Render Network’s founder, Jules Urbach, serves as OTOY’s strategic visionary and chief architect, pioneering computer graphics, streaming, and 3D rendering with over 25 years of industry experience. Since its founding in 2002, OTOY completed four funding rounds by 2020 and made a $50 million Series B investment in Light Field Labs in February 2023.
Since holding its initial ICO in October 2017, Render Network has evolved into an early but highly promising project. It conducted a private sale between January and May 2018. Currently, Render Network’s native token $RNDR has a market cap of $4.3 billion and an FDV of $6.1 billion, listed on major exchanges including Binance and OKX.
On November 2, 2023, the Render Foundation announced that Render Network successfully completed a core infrastructure migration from Ethereum to Solana, leveraging Solana’s high performance and low cost to further accelerate platform growth. As part of this transition, a new token, RENDER, was introduced and minted on Solana, while the legacy RNDR token remains valid. During the swap process, 1 RNDR equals 1 RENDER. Going forward, the Render Foundation will prioritize and fully support the RENDER token, marking a significant milestone in Render Network’s evolution.
NetMind.AI
NetMind.AI is a large-scale AI ecosystem composed of four core modules: compute market, chatbot, agent, and personal assistant. At its core, NetMind.Power is an Ethereum-based compute marketplace module providing GPU inference and training platforms. Since September 2023, NetMind.AI’s training/inference workload management platform has entered testnet phase. NetMind.Power supports a wide range of GPU models and integrates one-click import from Google Colab, offering a smooth user experience currently in free beta testing.
NetMind Chain is fully EVM-compatible. Its token $NMT has a market cap of $444M and an FDV of approximately $13B.
CUDOS
CUDOS, similar to Akash, is a delegated proof-of-stake (DPoS) blockchain built using Tendermint Core on the Cosmos SDK, featuring cross-chain functionality with Ethereum and Layer 2 solutions. Cudos Blockchain Compute aims to offer multi-chain EVM support and GPU computing capabilities.
$CUDOS has a market cap of $161M and an FDV of $275M.
Nuco.cloud
Nuco.cloud is a blockchain-based decentralized cloud computing service offering diversified solutions including AI training, data analytics, scientific research, and efficient rendering. Operating on both Ethereum and Telos blockchains, it leverages decentralization to deliver flexible and efficient computing resources. Users contribute their computing power to earn compute credits, while miners receive NUCO tokens as rewards. Token generation follows a proof-of-stake (PoS) and proof-of-research mechanism, supported by Eagle Capital GmbH. Notably, Nuco.cloud partners with CUDOS to jointly advance blockchain technology adoption.
To meet diverse user needs, Nuco.cloud offers multiple product lines—nuco.cloud SKYNET, nuco.cloud GO, nuco.cloud PRO, and CUSTOM—each tailored to specific scenarios. Utilizing advanced blockchain technology and intelligent pricing mechanisms, Nuco.cloud delivers a unique, efficient, and decentralized cloud computing platform.
Since its inception in 2017, Nuco.cloud launched its testnet in August 2023 and remains in testing phase. In its 2024 roadmap, Nuco.cloud plans a rebranding, release a B2B-focused PRO version, and seek listing on centralized exchanges (CEXs) to expand market reach.
In terms of compliance, Nuco.cloud meets EU GDPR standards and German regulations, distinguishing itself from competitors such as AWS, Azure, Google Cloud, and decentralized computing projects like GLM, Akash, RNDR, and IExec RLC. According to project materials (Deck), Nuco.cloud’s fees are significantly lower than other decentralized cloud providers—approximately 30% of comparable AWS configurations—demonstrating strong pricing competitiveness.
Currently, Nuco.cloud’s token $NCDT has a market cap of $44M and an FDV of $68M.
Dynex
Dynex is a neuromorphic supercomputing blockchain based on the DynexSolve chip algorithm, introducing a Proof-of-Useful-Work (PoUW) method to enhance speed and efficiency in decentralized networks. Dynex aims to provide computing power for AI, machine learning, fintech, biomedicine, and more, primarily utilizing GPU resources.
Launched in 2020 initially as a neuromorphic chip design project, Dynex later evolved into a blockchain, launching its mainnet in Q4 2022 and listing on Gate.io in Q1 2024.
$DNX was fairly launched in August 2022 with a deflationary model, having a market cap of approximately $86M and an FDV of $110M.
OctaSpace
OctaSpace is an open-source, scalable infrastructure of distributed cloud computing nodes, enabling access to distributed computing, data storage, services, and VPN. OctaSpace includes CPU and GPU compute, serving ML tasks, disk space, AI tools, image processing, and scene rendering using Blender.
Launched in 2022, OctaSpace operates on its own EVM-compatible Layer 1 blockchain. The blockchain uses a dual-chain system combining Proof-of-Work (PoW) and Proof-of-Authority (PoA) consensus. $OCTA has an FDV of about $73M and was fairly launched.
AIOZ Network
AIOZ Network is a Layer 1 decentralized computing platform for AI, storage, and streaming, interoperable with Ethereum and Cosmos. Supported by a decentralized content delivery network (dCDN), AI computing, and thousands of globally distributed nodes, users share computing resources for storing, transcoding, and streaming digital media, while supporting decentralized AI computing.
After more than six years of development, AIOZ launched its mainnet in December 2021. Infrastructure includes AIOZ dCDN, AIOZ IPFS, AIOZ W3S, AIOZ W3AI, and AIOZ Web3 Streaming, supporting Web3 storage, decentralized AI computing, live streaming, and video-on-demand (VOD).
According to its roadmap, Q1 will include AIOZ Transfer dApp, AIOZ multi-chain wallet browser extension, AlOZ Node V3, and AIOZ W3IPFS infrastructure; Q2 will deliver AIOZ W3AI; Q3 will launch AIOZ W3Stream and various AIOZ token standards; Q4 will implement AIOZ DEX.
Phoenix
Phoenix is a blockchain infrastructure platform for decentralized AI, compute scaling, and data-driven networks. Its products include AlphaNet—an AI platform for crypto trading—and NYBL, an AIGC metaverse project. Phoenix’s AI computing layer is a Web3-based infrastructure designed for scaling AI workloads and easily deploying AI-enabled applications, including deep learning, predictive analytics, LLMs, federated learning, and AI edge computing. Phoenix’s AI vertical solutions span multiple domains.
Recently, Phoenix partnered with Helium to launch the AI compute miner PhoenixNode, and its PhoenixLLM large language model service is now available on Telegram. Additionally, Phoenix will integrate with Telegram’s entire ecosystem.
Aethir
Aethir focuses on building scalable, decentralized cloud infrastructure for gaming and AI, based on Arbitrum. Launched in 2023, the project is currently in testnet phase. User experience closely resembles Web2 cloud gaming platforms, addressing testing, cross-platform, and distribution needs for AI and games. Key AI use cases include zero-latency inference, language interaction, and visual interaction. Checker node licenses had a whitelist sale from March 18 to 19 at 10:00 AM UTC, with public sales starting March 20 at 10:00 AM UTC, priced at 0.1259 ETH per node, incrementally increasing, and tradable as ERC721s on secondary markets. Checker nodes can be installed and run in Q2 2024, with token launch and mainnet activation expected in Q2. 15% of $ATH will reward Checker node operators.
Iagon
Iagon is a decentralized marketplace for storage and computing power based on Cardano. It released its testnet in 2023 and is open source.
OpFlow
OpFlow is a decentralized compute cloud platform focusing on AI and rendering. OpFlow Hosting is a decentralized cloud service compatible with Telegram bots. Its rendering service exclusively uses NVIDIA L40 GPUs. According to its roadmap, Phase One goals include: launching the token on Uniswap; activating OpFlow Hosting (Telegram Bot); and launching nodes on ETH, Linea, Eigen Layer, and other networks.
$OpFlow is an ERC20 token; further details have not been disclosed.
OpSec
OpSec is a suite of decentralized cloud computing solutions aiming to build the next-generation supercomputer. Solutions include Cloudverse, Cloudsec, OpSec Network, one-click node setup, DePIN hardware, and managed services.
The project is new, with no testnet yet released. $OPSEC is currently listed on DEXs.
Compute Market
Compute markets leverage decentralized mechanisms, allowing users to contribute their GPU and CPU resources to compute leasing or network training initiatives. While theoretically not reducing AI training costs or improving efficiency, these markets mobilize vast GPU resources quickly due to their openness and accessibility, providing robust computational support. This approach unlocks substantial asset backing and enables leveraged tokenization of GPU assets, creating new value and significance for compute markets.
However, such projects are often seen as “more hype than substance,” where core value lies less in actual technical utility and more in decentralization narratives (e.g., progression from traditional AI to decentralized inference, then to decentralized training). In practice, business models revolve largely around these narratives.
Technically, compute market projects capable of decentralized GPU operation often combine with the DePin concept. The combination of compute markets plus DePin, exemplified by io.net and Nosana, along with high-performance chains like Solana, demonstrates significant growth potential. Therefore, we recommend investors closely monitor early-stage projects combining Solana and AI, which could drive technological innovation and become investment hotspots.
From an investment perspective, early participation in GPU compute markets—especially during periods of FOMO—can yield high returns. Such participation not only generates incentives but also offers high input-output ratios, delivering substantial economic benefits.
Clore.ai
Clore.ai is a GPU compute rental platform built on PoW. Users can rent out their GPUs for AI training, rendering, mining, and other tasks. It uses an optional Proof-of-Holding (PoH) mechanism: the more tokens suppliers hold, the more CLORE they receive.
$CLORE has a market cap of $90M and an FDV of $110M. Each block allocates 50% to miners, 40% to custodians, and 10% to the team, with a hard cap of 1.3 billion tokens.
Nosana
Nosana is a GPU-focused cloud computing provider on Solana, with open-source code. Founded in 2021, the project currently has no disclosed fundraising. Co-founder Sjoerd Dijkstra has extensive DevOps experience; Jesse Eisses has solid technical/ML background; Laurens Verspeek is technically skilled with rich development experience. In mid-October 2023, Nosana announced a pivot from CI/CD use cases to AI inference. The platform remains in early development, with minimal data visible on nosana explorer—currently reporting 94 GPU nodes and 160,000 completed inferences, requiring further investigation.
Llama 2 and Stable Diffusion AI inference workloads will be integrated into the network. The testnet has three phases; currently in Phase One. According to official Twitter data, over 1,000 devices have registered for Phase Two. Revenue and data models remain undisclosed. Plans include mainnet launch in the first half, enhancing consumer-grade node support; in the second half, establishing community connector libraries and official connectors for PyTorch, HuggingFace, TensorFlow, etc.
$NOS launched in January 2022, currently with a market cap of $510M and an FDV of $608M.
io.net
io.net is an innovative decentralized AI compute marketplace built on Solana blockchain technology, currently in testnet phase. The platform aims to provide access to cloud computing resources at lower costs than traditional centralized services, achieving a valuation of $500 million. io.net’s core appeal lies in supporting various AI computing needs—batch inference, parallel training, hyperparameter tuning, reinforcement learning—with a backend infrastructure composed of modular layers enabling effective resource management and automated pricing. Currently, it offers NVIDIA RTX 3090 computing resources at $0.20 per hour.
Founded by Ahmad Shadid, who built a GPU computing network in 2020 for Dark Tick, a machine learning quantitative trading firm. Regarding funding, io.net has received lead investment from Hack VC, with additional support from prominent investors including Multicoin Capital, Solana Labs, Aptos Labs, and notable figures such as Solana’s founder Anatoly Yakovenko.
For participation, io.net offers diverse ways to engage, including mining income and staking rewards. From March 1 to April 28, 2023, the platform ran its first points incentive program, expecting to issue its $IO token in April. Mining rewards are calculated based on detected node metrics: internet bandwidth, memory, CPU clock speed, and floating-point operations. Users can also earn yields by staking IO Coin, receiving 1–3% of all rewards earned by participants.
Notably, io.net’s team modeled its points calculation after RNDR’s airdrop rules, emphasizing bandwidth importance in reward allocation. Analysis shows high-bandwidth devices gain clear advantages in final reward distribution, suggesting participants should prioritize improving device bandwidth when joining the io.net network.
Gensyn
Gensyn is a Layer 1 protocol focused on training deep learning models, aiming to improve efficiency and accessibility through innovative design and technical solutions. Since launch, Gensyn has completed two funding rounds: a $6.5 million seed round led by EdenBlock in March 2022, and a $43 million Series A led by a16z in June 2023.
Gensyn’s technical core lies in its verification system and compute supply framework, along with an efficient solution stack. The verification system consists of four key roles: Submitters, Solvers, Verifiers, and Whistleblowers, collectively ensuring a trustless, i.e., non-custodial, distributed verification mechanism. Moreover, Gensyn strives to make its compute network accessible to various devices, including consumer-grade gaming GPUs and former ETH PoW mining GPUs. Its efficient machine learning training solution leverages probabilistic proofs of learning, graph-based precise localization protocols, and Truebit-style incentive games, aiming to drastically reduce training costs and improve efficiency.
In terms of cost, Gensyn shows strong competitiveness. Its average hourly machine learning training cost is approximately $0.40—lower than Akash’s $1.50 per hour GPU usage fee and within io.net’s $0.10–$1.00 range—highlighting Gensyn’s strengths in cost optimization.
Gensyn’s code is not yet open source, and there is no current token issuance plan. Development will proceed in three phases: testnet, canary net on Kusama relay chain, and mainnet on Polkadot relay chain. Currently, the project remains in testnet phase. Over the next 3–6 months, Gensyn plans to continue technical optimization and prepare for the next development stage. Its vision is to become the foundational layer for machine learning computation—akin to Ethereum for smart contract execution—indicating Gensyn’s long-term planning for the convergence of machine learning and blockchain applications.
Nimble
Nimble's core vision is to build a decentralized AI ecosystem encompassing three markets: data, compute, and builders. This design aims to lower the barrier for training AI models, enabling broader participation in AI innovation. Positioned as a Composable AI Project, Nimble emphasizes the composability of various AI resources and components within a decentralized framework.
Nimble mainnet launched on March 21, 2024, and was selected into BinanceLabs’ seventh season Most Valuable Builder (MVB) Incubator Program on March 1. No complete security audit report is publicly available. Nimble’s valuation is $30 million, with $6 million raised in funding.
Mainnet went live last weekend, allowing users to mine by contributing GPU power. Over 1,500 GPUs have already joined the network.
Morpheus AI
Morpheus AI is a decentralized compute marketplace. Built on Arbitrum, it helps users create agent AIs based on large language models, which then invoke smart contracts. This design expands the functional potential of decentralized applications (DApps) and creates incentive mechanisms for compute providers, stETH stakers, protocol contributors, and community API runners. Within the Morpheus AI ecosystem, users can earn MOR tokens via staking or code submissions.
Kuzco
Kuzco is a distributed GPU cluster on Solana for LLM inference. Though newly launched, it progresses rapidly, having released a public beta for Mac and Linux users, with over 1,400 online working nodes. Kuzco provides an efficient environment and code support, enabling users to pre-install required models and environments locally, completing tasks via network transmission alone. Compared to ionet, it achieves better resource utilization, reaching up to 90% single-GPU usage.
Currently, contributors receive $KZO points as rewards. Although expected returns, roadmap, and token economics remain unclear, Kuzco has attracted attention from Solana’s official team. No fundraising has occurred yet, but token issuance may happen soon.
Golem
Golem is a decentralized compute marketplace based on Ethereum, focusing since 2015 on providing peer-to-peer CPU compute sharing services via Ethereum. One of the earliest peer-to-peer compute network protocols, Golem has established a network node system comprising providers and requesters. Starting in 2022, Golem expanded into the GPU market, launching Phase One of its GPU Beta Testing Program and initiating Phase Two in March 2024.
$GLM has a total supply of 1 billion, all currently in circulation, with a market cap of $561 million. The codebase is open source.
Node AI
Node AI is a decentralized GPU compute marketplace, with parent company EyePerformance. Rental rates: ~$0.85/h for A100, ~$0.22/h for A10. $GPU currently has a market cap of $121M.
GPU.Net
GPU.Net is a decentralized GPU resource network aiming to meet compute demands for generative AI, Web3 metaverse, cryptocurrency mining, and high-end graphics rendering through decentralized GPU infrastructure. In August 2023, GPU.Net raised $500,000 in funding from Momentum 6 and Alphablockz, with support from NVIDIA and Taanga Studios. Early network pre-registration is now live.
GamerHash
GamerHash utilizes unused computing power from gamers’ PCs during high-spec gameplay to mine cryptocurrencies. For example, if a player uses only 15% of their PC’s capacity, GamerHash automatically harnesses the remaining unused power for cryptocurrency mining without requiring additional user action or monitoring. To serve users with lower-end hardware, GamerHash introduced a Play&Earn feature. Users can complete specific tasks—such as playing certain games or downloading apps—to earn GUSD, an internal currency pegged to the US dollar, creating a new income stream for users with limited hardware capabilities.
NodeSynapse
NodeSynapse is a GPU compute marketplace offering server hosting, Web3 infrastructure, GPU computing, and a unique revenue-sharing model for token holders.
Model Assetization and Model Training
Within the convergence of blockchain and AI, a key development direction involves assetizing AI models or running business models via tokenization. This approach is gradually demonstrating value across various applications such as security auditing, chatbots, and AI advisors. Due to relatively low technical barriers and implementation difficulty, the landscape includes both well-resourced top-tier projects and numerous early-stage ventures. Thus, distinguishing project quality and investment merit becomes crucial in blockchain-AI integration efforts.
The key lies in identifying whether genuine innovation and technical moats exist. Simply trading ownership or usage rights of AI models does not constitute true technological advancement. Real breakthroughs should focus on effectively verifying model outputs and ensuring reliable model execution in decentralized environments.
Saharalabs
SaharaLabs aims to address data sharing and privacy challenges in AI model training through its two core products: Knowledge Agent and Data. The project provides semi-automated autonomous agents and data operations tools for individuals and enterprises, promoting data sharing and decision-making while ensuring privacy security. SaharaLabs has already attracted 30 enterprise clients and achieved positive cash flow, demonstrating market viability and practical utility. Led by Professor Sean and Tyler, the team exceeds 30 members. Professor Sean teaches at the University of Southern California and has received multiple honors, including Samsung’s AI Researcher of the Year. Tyler has extensive blockchain industry experience, formerly serving as Investment Director at Binance Labs.
Its Knowledge Agent product offers customizable semi-automated agents that support sharing and processing external or internal data for decision-making. These agents can be fine-tuned on private data, include privacy protection and data provenance verification tools, and provide a developer toolkit for multi-agent collaboration. The Sahara Data Marketplace offers a comprehensive data operations suite including data collection, labeling, quality assurance (QA), and project management, solving data privacy issues and already adopted by renowned institutions such as MIT and Microsoft.
SaharaLabs has raised $600 million in funding and secured $6 million in seed funding led by Polychain Capital, with participation from Sequoia Capital, Samsung Next, and Matrix Partners. The project roadmap includes: launching the Sahara Data marketplace in Q2 2024; testnet and Sahara ID in Q3; and mainnet and Sahara Agent in Q4.
Bittensor
Bittensor is a decentralized open-source project aiming to create a neural network protocol on blockchain. This protocol enables AI-driven decentralized applications (dApps) and facilitates peer-to-peer value exchange between AI models. Bittensor was co-founded by Jacob Robert Steeves, a former Google software engineer, and Ala Shaabana, a former assistant professor at the University of Toronto and postdoctoral fellow at the University of Waterloo. James Woodman serves as Chief Operating Officer. The project underwent a "fair launch" in 2021 with no pre-mining, VC rounds, or private sales—the TAO token was mined entirely by miners. Known investors and market makers include DCG, GSR, Polychain Capital, and Firstmask.
The Bittensor network is primarily composed of miners and validators. Miners submit pre-trained models to earn rewards, while validators ensure output validity and accuracy, selecting the best responses to return to users. For example, when a user requests an AI chatbot response, the query is sent to validators, who forward it to miners, rank their outputs, and return the highest-ranked result—regardless of how many nodes operate on the Bittensor network.
Key development areas for Bittensor include:
-
Subnets: All AI applications on Bittensor run on subnets—32 in total, each with specialized use cases. Entry threshold is 6,000 $TAO, costing over $4 million. Miners and validators earn 7,200 $TAO daily; if a subnet performs well, daily earnings can reach hundreds of thousands of dollars; otherwise, the subnet risks elimination.
-
Miners/Validators: The barrier to becoming a miner or validator is high. Miners must pass hardware performance tests, while validators require large $TAO stakes and continuous uptime. Miners failing to provide optimized or unique large LLMs will be quickly removed from the network.
The TAO token has a total supply of 21,000,000, all in circulation, with a market cap of $4,474,463,212. Listed on MEXC, Gate.io, KuCoin, and Bitget. Retail investors can stake $TAO to existing validator nodes, earning an approximate annual yield of 15.85%.
iExec RLC
iExec RLC is a decentralized cloud resource platform based on Ethereum, aiming to transform current cloud computing models through integrated DApps, compute markets, and data markets. The platform uses its proprietary Proof-of-Contribution (PoCo) consensus mechanism to ensure off-chain computation integrity and security while protecting against dishonest participants. iExec was co-founded by Dr. Gilles Fedak and Dr. Haiwu He, with Fedak serving as CEO. Before blockchain entrepreneurship, Fedak achieved groundbreaking results in cloud computing and large-scale parallel systems, winning several best paper awards. Dr. Haiwu He, leading APAC operations, was a “Hundred Talents” professor at the Chinese Academy of Sciences, directing research in high-performance computing (HPC) and cloud computing. In 2017, iExec raised $12 million through an ICO.
iExec’s core functionality allows users to run containerized applications, mainly targeting short-duration jobs and one-off tasks. While not supporting broad use cases like Web or API-based applications, iExec has successfully powered DApps on its sidechain and established worker pools sorted by task size. Its compute resource bidding system allows users and apps to publish work orders based on preset criteria, while cloud providers (workers) bid their compute capacity. iExec also introduces the pool concept to organize workers, which can be public or private, resembling traditional cloud providers.
The PoCo consensus algorithm acts as a defense mechanism for the iExec market, guaranteeing trusted execution on decentralized networks and preventing adverse outcomes and unreasonable disputes. To date, $RLC has a market cap of $322 million and an FDV of $387 million. Since launching in 2016, iExec has achieved notable success in decentralized cloud computing and became a key member of Intel’s AI ecosystem in 2020.
iExec continues to advance its technology and service offerings, with future plans including: research and development (R&D) for AI and large language models (LLMs); R&D for Privacy-Enhancing Technologies (PET); development of the iExec Oracle IDE; Privacy Pass campaign; second round issuance of Worker Pass; launch of the Dapp Store web interface; third season of Questbook bounties; third and fourth seasons of Exec Web3 incubator; and upgrades to the PoCo consensus algorithm.
Allora
Allora is a self-improving, decentralized AI network that achieves more accurate inference through a prediction market mechanism. AI/ML agents on the Allora network broadcast predictions on a peer-to-peer network using their data and algorithms, referencing each other’s forecasts to evaluate accuracy. The network’s consensus mechanism combines these predictions and evaluations, allocating rewards based on the quality of each agent’s forecasting and assessment. This carefully designed incentive structure enables Allora to continuously learn and improve, adapting as the market evolves.
Built on Cosmos as an independent chain with PoS, Allora rewards validators and AI worker nodes. Nodes perform inference while mutually verifying the accuracy of each other’s results, directly evaluating inference quality across agents and distributing rewards accordingly.
Allora’s internal testnet Edgenet launched on March 5, currently collaborating with multiple projects and institutions to run test nodes, with mainnet planned for Q2. Its investor roster is impressive, including Polychain, Framework, Blockchain Capital, CoinFund, Delphi Ventures, dao5, and Distributed Global.
lPAAL AI
lPAAL AI is an AI ecosystem built using custom data sources and LLMs, allowing users to create personalized AIs capable of professional strategy trading, market intelligence gathering, and deployment across platforms like Telegram and Discord. PAAL AI tools include MyPaal, AutoPaal, and AutoPaal X. Backed by investments from Google Cloud, Coingecko, and OKX. PAAL has a total supply of 1 billion, with approximately 650 million in circulation, and a market cap of about $100 million.
MyShell
MyShell is an AI platform based on opBNB that enables users to create chatbots, allowing creators to call different models and external APIs uniformly, and enabling third-party model and API providers to integrate their services for developers. Shell Points and Shell Coin incentivize creator and user engagement (Shell Coin serves as an airdrop voucher; points can be redeemed for Coins). Raised $5.6 million in seed funding in October 2023 led by INCE Capital, with participation from Folius Ventures, Hashkey Capital, SevenX Ventures, TSVC, and OP Crypto. Binance Labs officially announced MyShell’s inclusion in its sixth incubation batch. Valued at approximately $57 million. Currently in testnet development phase.
Qubic
Qubic is a Layer 1 PoW platform using PoW compute power for AI training. Mainnet launched in April 2022. Founded by Sergey Ivano (also known as Come From Beyond or CFB). CFB was the third person in history to mine Bitcoin, founded NXT—one of the most successful ICOs—and is also the founder of Iota.
-
Quorum Protocol: Processes transactions and executes smart contracts via 676 computers (validators), ensuring finality and reliability.
-
AI Aigarth: An AI software running atop Qubic, drawing information from tens of thousands of Qubic-AI miners whose computing power generates billions of artificial neural networks.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
Add to FavoritesShare to Social Media
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@TrustlessLabs
