Subnets

Chutes delivers serverless AI compute enabling users to deploy, run and scale any AI model in seconds with zero infrastructure management. The platform provides instant model deployment with automatic load balancing and scaling across distributed GPU resources. Miners provide compute capacity while validators ensure quality and availability of the inference services.

Lium.io by Datura AI provides distributed computing and storage infrastructure, enabling scalable processing and data management for AI applications. The subnet combines compute resources with decentralized storage capabilities, offering a complete infrastructure layer. Developers access resources through APIs with automatic scaling based on demand.

Targon is a next-generation AI infrastructure platform leveraging Confidential Compute (CC) and Protected PCIe (PPCIE) technology to secure the entire stack from hardware to application. By providing trusted execution environments with hardware-enforced memory encryption and GPU TEE isolation, Targon enables verifiable and trustworthy AI inference operations. The platform supports NVIDIA Confidential Compute with AMD SEV-SNP integration on the roadmap.

Vanta Network is a quantitative trading subnet delivering ML-powered signals from the world's best traders and quant systems across Forex and cryptocurrency markets. Miners compete by providing highly efficient long/short futures signals that generate maximum returns while staying under strict 10% drawdown limits. Only elite traders with consistent performance survive the competitive elimination system based on plagiarism detection, drawdown monitoring, and probation periods.

τemplar is a decentralized training framework enabling large-scale AI model training across heterogeneous compute resources over the internet. By connecting diverse computational nodes through carefully designed incentive mechanisms, it makes collaborative model training possible while ensuring honest participation. The system coordinates miners who train on data subsets with validators who evaluate contribution quality and assign rewards accordingly.

Ridges enables distributed edge computing, processing data close to its source for reduced latency and bandwidth optimization. The subnet connects edge devices and IoT sensors to decentralized processing nodes, enabling real-time analytics without centralized cloud dependencies. Miners provide edge compute resources while validators ensure processing quality.

IOTA (Incentivized Orchestrated Training Architecture) is a groundbreaking framework for pretraining large language models across heterogeneous, unreliable, and permissionless machines. Using data- and pipeline-parallel architecture, it accelerates training while reducing hardware requirements for participants. The system enables miners to compete in processing activations while validators verify work through spot-checks.

Gradients on Demand (G.O.D) runs competitive tournaments where miners' open-source AutoML scripts train LLMs and diffusion models on dedicated validator infrastructure. Tournaments last 4-7 days with 72-hour gaps between events, offering exponentially higher weight potential for top performers. First-place winning scripts are released to github.com/gradients-opensource, advancing open-source AI development.

BitMind is the first decentralized AI-generated content detection network on Bittensor, built to identify synthetic media like deepfakes. It rewards accurate detection algorithms through a competitive, incentive-driven system of validators and miners. The subnet promotes digital trust with open-source tools, benchmarks, and APIs for broad integration.

Subnet 5 aims to develop the best-performing, most general-purpose text embedding model in the world. The model's performance will be evaluated against an infinitely large and dynamic dataset, serving as a proxy for an infinitely generalized benchmark, to ensure the highest possible level of domain generalization. As a network, our model will continuously improve and adapt to the latest real-world knowledge. This dynamic evaluation will ensure that SN5’s embedding model not only surpasses existing state-of-the-art (SOTA) models, pushing the boundaries of industry performance, but also remains consistently competitive.

The Next Generation of Computer Vision Teaching machines to understand our visual world autonomously, starting with the dynamic complexity of soccer - where every movement, every play, and every moment trains our AI to decode the world around us.

Hippius is a comprehensive decentralized infrastructure subnet providing distributed storage, compute, and blockchain management capabilities on Bittensor. It leverages IPFS integration for file storage while offering virtualized compute resources through hypervisor-based miners. The subnet includes reputation-based worker management, marketplace functionality, and advanced authorization systems.

NOVA focuses on next-generation AI architectures and novel approaches to machine learning, serving as an innovation testbed for emerging techniques. The subnet encourages experimentation with new model designs, training methods, and inference optimizations. Miners implement cutting-edge research while validators benchmark against state-of-the-art baselines.

Nineteen.ai provides fast, reliable AI model inference through a decentralized network of miners serving model requests. The subnet optimizes for low latency and high throughput, with validators ensuring inference quality and response accuracy. Developers can integrate via API to access distributed AI computing power without managing their own infrastructure.

404-GEN democratizes 3D content creation through AI-powered asset generation for games, AR/VR/XR experiences, and virtual worlds. The subnet uses multiple LLM-powered prompt generators to create diverse, realistic prompts for 3D model generation, preventing memorization through continuous variety. The vision enables AI-native games where all assets, dialogue, and sound are generated at runtime from simple prompts.

Coldint (COLaborative Destributed INcentivized Training) advances collaborative, distributed model training and research with incentives for publishing improvements. Forked from subnet 9 pretraining, it addresses the lack of incentive for small incremental improvements. Initial competition rewards miners for improving model scores on the Fineweb-edu-2 dataset, with additional rewards for code contributions and key insights.

Reimagining Bitcoin Mining Traditional pools are built for profit, offering no transparency on the code and how rewards are distributed. TAOHash, powered by Bittensor Subnet 14, changes that by putting miners first with a transparent, open-source platform, and a unique ownership model.

Apex is the pioneering decentralized AI inference subnet on Bittensor, achieving deep-researcher reasoning capabilities as the first agent on the protocol. As Subnet 1, it represents the most intelligent inference model on the network with advanced multi-step reasoning. The subnet pioneered incentivized intelligence on Bittensor and continues to set the standard for decentralized AI inference quality.

Dippy Studio is an enterprise-grade generative media platform powering fast, deterministic image inference for Dippy's 7M+ AI companion app users. The subnet leverages a custom in-house Dippy Inference Engine optimized for speed and cost efficiency, delivering the fastest and cheapest inference available. Phase roadmap expands to audio, video, 3D generation while maintaining production-ready reliability from day one.

Quantum Innovate explores advanced quantum computing applications including hybrid quantum-classical algorithms and quantum machine learning integration. The subnet provides a research platform for developing quantum solutions to optimization, cryptography, and simulation problems. Miners contribute quantum computing resources while validators benchmark algorithm performance.

Synth generates the world's most powerful synthetic price data for probabilistic forecasting, capturing full distributions of possible price movements rather than single-point predictions. Miners produce 1000 simulated price paths per asset (BTC, ETH, SOL, XAU) at 5-minute intervals over 24-hour horizons. Scoring uses Continuous Ranked Probability Score (CRPS) measuring both calibration and sharpness against actual price movements.

Basilica creates a trustless marketplace for GPU compute through hardware verification, SSH-based remote validation, and smart HTTP gateway for load-balanced network access. The validator verifies hardware capabilities and maintains GPU profiles while miners manage executor fleets handling assignments. Features include binary validation system, collateral mechanisms, and one-line installation via curl.

Graphite is a decentralized network for solving graph optimization problems, currently specializing in the Traveling Salesman Problem (TSP)—a classic challenge where complexity grows exponentially with nodes. Miners compete to find optimal routes using their own algorithms while validators generate synthetic problems and score solutions. The platform supports both coordinate inputs for undirected graphs and edge inputs for directed graphs.

Quantum Compute explores quantum computing applications within the Bittensor network, researching quantum algorithms and hybrid classical-quantum approaches. The subnet provides a platform for quantum computing researchers to experiment with decentralized quantum resources. Focus areas include quantum optimization, cryptography research, and quantum machine learning integration.

Safe Scan provides comprehensive security scanning services including vulnerability assessment, malware detection, and threat analysis. The subnet offers sandboxed analysis environments for safely examining suspicious files and URLs. Miners run security scans while validators verify detection accuracy against known threat databases.

Grail provides premium-tier AI services with curated, high-quality model access and enterprise-grade reliability guarantees. The subnet focuses on quality over quantity, offering vetted models with consistent performance and dedicated support. Miners meet strict quality standards while validators ensure service level compliance.

BitAds is a decentralized advertising network built on Bittensor, enabling permissionless ad placement and distribution without traditional gatekeepers. Miners earn rewards by serving and distributing advertisements across the network while validators ensure ad quality and delivery metrics. The platform brings Web3 advertising infrastructure to the Bittensor ecosystem with transparent, verifiable campaign performance.

Bitcast provides AI-powered audio and podcast services including content creation, transcription, and streaming. The subnet enables podcast generation, voice synthesis, and audio analysis at scale. Miners process audio workloads while validators ensure content quality and authenticity.

We incentivize the prediction of future events. The prediction space is based on binary future events such as the ones listed on Polymarket. We are always actively expanding to new data providers. Most of this data is then processed by an LLM pipeline which handles the event generation. We are focused on judgemental forecasting rather than statistical forecasting. We hence expect the models used by miners to be LLMs.

ComputeHorde transforms untrusted GPU resources from miners into trusted, verified compute power for Bittensor validators across all subnets. Through collateral-based trust mechanisms, commit-reveal weight protection, and hardware verification, it creates a decentralized marketplace where validators can access scalable GPU computing without relying on centralized cloud providers. The subnet supports A6000 and A100 hardware classes with SDK integration for easy adoption.

Leadpoet combines AI-powered lead generation with content creation for marketing automation. The subnet helps businesses discover potential customers through intelligent prospecting while generating personalized outreach content. Miners develop lead scoring models and content generators while validators evaluate lead quality and content effectiveness.

Bitsec.ai provides AI-powered cybersecurity services including threat detection, vulnerability scanning, and security analysis. The subnet leverages machine learning to identify malware, suspicious patterns, and security weaknesses across systems and code. Miners process security scans while validators verify detection accuracy against known threats.

Developing state of the art intelligence depends disproportionately on the underlying datasets. SN13 provides the Bittensor ecosystem with a diverse, high quality dataset from a diverse array of sources.

Bettensor is a sports prediction subnet where fans and ML researchers compete to predict outcomes of their favorite sporting events. Miners receive upcoming games with odds, submit simulated wagers, and are rewarded for correct predictions. The platform supports Baseball, Football (Soccer), and American Football with minimal compute requirements—human miners just need to consistently submit predictions.

Liquidity subnet enables decentralized market making and liquidity provision across DeFi protocols. The subnet coordinates liquidity providers to ensure deep, stable markets for token trading with optimized capital efficiency. Miners provide liquidity while validators monitor pool health and distribution efficiency.

ForeverMoney (九八) focuses on long-term value preservation and sustainable wealth management strategies. The subnet applies AI to dynasty wealth planning, intergenerational asset transfer, and capital preservation techniques. Miners develop preservation models while validators evaluate strategy effectiveness over extended timeframes.

Dojo transforms conventional GANs into a competitive, decentralized zero-sum game where miners act as both generators and discriminators. Unlike traditional GANs mimicking ground truth, Dojo challenges miners to create outputs superior to high-quality baselines. Mining requires no server setup—just register, load your hotkey to a browser wallet like Talisman, and receive tasks at dojo.network.

EfficientFrontier by SignalPlus identifies optimal risk-weighted crypto trading strategies through live trading performance data. Strategies are ranked on the "impossible trinity" of high returns, minimal drawdowns, and fast capital recovery. Scoring uses exponentially weighted daily returns against rolling max drawdowns, with size adjustments rewarding larger portfolios and penalties for excessive leverage above 50%.

MetaHash provides cryptographic content hashing and verification for media authenticity, combating deepfakes and misinformation. The subnet creates verifiable fingerprints for images, videos, and audio, enabling detection of manipulated content. Miners process media while validators maintain hash databases and verify integrity claims.

AlphaCore provides foundational AI processing capabilities including model inference, training support, and optimization services for the Bittensor ecosystem. The subnet serves as core infrastructure enabling other applications to leverage distributed AI computing power. Miners offer processing resources while validators benchmark performance and quality.

τensorprox operates as a tensor proxy service providing optimized routing, load balancing, and caching for AI workloads. The subnet acts as an intelligent gateway distributing requests across available compute resources while maintaining response quality. Miners operate proxy nodes while validators monitor routing efficiency.

Desearch delivers AI-powered decentralized search with real-time access to metadata from X, Reddit, Arxiv, and the broader web. Unlike centralized search engines, it provides unbiased, verifiable results through community-driven optimization by miners, validators, and users. Features include sentiment analysis determining emotional tone of social posts and comprehensive metadata analysis for research applications.

τenex operates as a tensor exchange marketplace enabling trading of AI models, datasets, and computational resources within Bittensor. The subnet facilitates discovery and exchange of AI assets with tokenized transactions and transparent pricing. Miners list and serve assets while validators ensure quality and authenticity.

Autoppia's Web Agents subnet uses the Infinite Web Arena (IWA) benchmark to incentivize development of autonomous web agents that navigate and complete tasks on dynamically generated websites. Unlike static benchmarks that can be gamed through memorization, IWA uses generative AI to create continuous novel web challenges. The permissionless Autoppia Web Operator enables business automation that adapts to interface changes.

InfiniteHash provides scalable cryptographic hashing infrastructure capable of processing massive throughput requirements. The subnet supports applications needing high-volume hash computation including blockchain validation, content fingerprinting, and security operations. Miners contribute hashing power while validators benchmark throughput and accuracy.

Kinitro incentivizes the emergence of embodied AI agents that can conquer various tasks across different environments through real-time competition. Miners publish agents to compete while validators perform rollouts and evaluate performance through standardized benchmarks. The subnet features a comprehensive dashboard for viewing agent performance and a robust evaluation cluster powered by Ray and Kubernetes.

CheckerChain provides blockchain analytics and transaction verification services for compliance and security applications. The subnet traces fund flows, identifies suspicious patterns, and generates compliance reports across multiple chains. Miners process chain data while validators ensure analysis accuracy against known patterns.

Gittensor integrates AI with Git-based development workflows, providing code analysis, automated review, and collaboration assistance. The subnet analyzes repositories to understand codebases, suggest improvements, and accelerate development. Miners process code analysis while validators evaluate suggestion quality and accuracy.

Vidaio specializes in AI-powered video understanding, generation, and processing. The subnet enables video content analysis, automatic editing, and AI video creation from text or image prompts. Miners process video workloads while validators evaluate output quality and ensure generation meets content guidelines.

HappyAI focuses on positivity, mood enhancement, and spreading joy through AI-generated content and interactions. The subnet creates uplifting content, positive affirmations, and happiness-boosting experiences. Miners generate positive content while validators ensure genuineness and avoid toxic positivity.

TAO Private Network provides secure, encrypted communication channels for privacy-preserving data transfer within the Bittensor ecosystem. The subnet offers end-to-end encryption, anonymous routing options, and secure messaging capabilities. Miners operate relay nodes while validators ensure network integrity and routing efficiency.

Gaia focuses on geospatial AI and Earth observation, processing satellite imagery and environmental data for climate analysis, mapping, and location intelligence. The subnet leverages remote sensing data to provide insights on land use, vegetation, urban development, and environmental changes. Miners process imagery while validators ensure analysis accuracy.

Agent Builder enables creation and deployment of autonomous AI agents without requiring extensive coding knowledge. The subnet provides visual tools and templates for building agents that can perform tasks, make decisions, and interact with external systems. Miners host deployed agents while validators ensure safe and reliable execution.

CliqueAI specializes in social network analytics and community intelligence, extracting insights from social graphs and community interactions. The subnet analyzes social dynamics, identifies influential members, and detects emerging trends across platforms. Miners process social data while validators ensure analysis accuracy and privacy compliance.

Reboot focuses on network restart and recovery capabilities, ensuring Bittensor ecosystem resilience during disruptions. The subnet coordinates synchronized restart protocols and monitors network health to maintain continuity. Validators track system status while miners provide recovery infrastructure for graceful failure handling.

Investing subnet provides AI-powered investment analysis including market research, portfolio optimization, and risk assessment. The subnet combines quantitative analysis with AI insights to help investors make data-driven decisions. Miners develop analysis models while validators evaluate prediction accuracy and portfolio performance.

Neza provides a general-purpose AI assistant subnet delivering versatile task completion capabilities across diverse domains. The subnet handles a wide range of requests from information queries to creative tasks with natural language interaction. Miners compete on response quality while validators evaluate helpfulness and accuracy.

Chunking subnet optimizes document segmentation for RAG (Retrieval Augmented Generation) pipelines and embedding workflows. Intelligent chunking preserves semantic boundaries and context, improving retrieval quality for vector databases. Miners compete to produce the best text segmentation strategies while validators evaluate chunk quality against retrieval performance metrics.

Distributed Training subnet enables decentralized training of large language models using the hivemind library for peer-to-peer gradient sharing and all-reduce operations. Miners contribute GPU resources to collaborative training runs, with model checkpoints stored on Cloudflare R2. The subnet supports multi-GPU configurations per miner with automatic updates and WandB integration for monitoring training progress.

This subnet slot is parked and reserved for future development while currently providing support infrastructure for Dippy Studio (SN11) operations. The parking allows for strategic resource allocation and ensures the slot remains available for high-priority initiatives. When activated, it will expand the Dippy ecosystem with complementary capabilities.

Babelbit enables cross-language AI translation and multilingual communication, breaking down language barriers through decentralized translation services. The subnet supports real-time translation across multiple languages while preserving context, nuance, and cultural meaning. Miners compete on translation quality with validators evaluating accuracy against reference translations.

BitAgent revolutionizes task automation by merging large language models with your favorite apps including web browsers, Discord, and other Bittensor subnets. Miners compete on the Berkeley Function Calling Leaderboard (BFCL) to develop the best tool-calling language models without requiring API subscriptions or heavy compute. The subnet powers downstream applications like GoGoAgent and MSP Tech for real-world intelligent agency.

CANDLES is a decentralized cryptocurrency candle prediction network where miners compete to forecast price movements—both candle colors (direction) and closing values—for hourly, daily, and weekly timeframes. Validators score predictions against actual market data, rewarding accuracy in both direction and price proximity. Web interface at candlestao.com enables easy prediction file generation.

The Finetuning subnet 37 rewards miners for fine-tuning Large Language Models (LLMs). At launch the first competition is evaluated with data generated from a continuous stream of synthetic data provided by subnet 18. It is a continuous fine-tuning benchmark, with new data generated daily.

FLock.io seeks to decentralise training and value alignment. We ensure that AI objectives match the public’s ethics and societal aims, that decision-making falls to communities, and that usefulness is a top priority. FLock knocks down barriers impeding participation in the ecosystem. We allow developers to provide models, data, or compute in a modular way. The result: a plethora of fit-for-purpose models created by, for, and under the stewardship of the communities.

ChipForge (Tatsu) focuses on AI hardware design and chip optimization, developing custom ASIC designs and acceleration strategies for AI workloads. The subnet brings together hardware engineers and AI researchers to create power-efficient, high-performance computing solutions. Miners contribute design optimization while validators benchmark against performance targets.

It's AI is the world's most accurate AI text detector, achieving 94.9% accuracy at 5% FPR on the RAID benchmark—outperforming commercial solutions like GPTZero, Winston, and Originality. Beyond classification, it provides deep scan functionality showing AI probability for each sentence. Available as web app, Chrome extension for instant highlighting detection, and API with Britishiser to convert American English LLM output.

Swap subnet incentivizes miners to provide liquidity to the TAO/USDC pool on TaoFi, the native decentralized exchange for the Bittensor ecosystem. Miners are scored based on how much fees their liquidity positions have earned over the past 24 hours, creating a direct incentive to maintain deep, active liquidity. This brings essential DeFi infrastructure to Bittensor, enabling seamless token swaps and price discovery.

OMEGA Any-to-Any creates state-of-the-art multimodal models that jointly learn from all modalities—text, audio, video, images—building toward a unified understanding of reality. Current focus is voice-to-voice generation, training models that fluently listen, understand, and respond. The innovative 75/25 emissions split rewards models exceeding performance baselines, functioning as a self-sustaining open research lab.

CoinMetrics Precog serves as an arena to identify analysts with the best strategies for anticipating Bitcoin price movements, recruiting them to share insights with the broader community. High-frequency predictions every 5 minutes with hourly settlement enable continuous feedback that traditional derivatives markets cannot provide. Scoring rewards precise price level predictions and most likely trading bands.

BitQuant implements a decentralized AI framework for quantitative DeFi analysis, enabling natural language queries for ML-powered market analytics, portfolio optimization, and risk assessment. Powered by the BitQuant agent with capabilities including pool discovery, TVL analytics, price trend analysis, drawdown calculation, and wallet analytics. Users can analyze tokens, DeFi pools, and portfolio risk across multiple chains including Ethereum and Solana.

Yanez MIID (Machine Intelligence Identity Detection) provides AI-powered identity verification and fraud prevention for compliance applications. The subnet offers KYC automation, document verification, and fraud detection capabilities using machine learning models. Validators ensure detection accuracy while miners process verification requests at scale.

Zeus leverages advanced AI models to forecast environmental variables using ERA5 reanalysis data from the Copernicus Climate Data Store—the largest global environmental dataset with hourly measurements from 1940 to present. The subnet incentivizes development of data-driven algorithms that can predict weather and climate faster, cheaper, and potentially more accurately than traditional physics-based numerical weather prediction models.

Vericore provides verification and validation services ensuring data authenticity, integrity, and trustworthiness across the Bittensor ecosystem. The subnet offers cryptographic verification, source validation, and provenance tracking for AI-generated and human-created content. Miners process verification requests while validators maintain trust standards.

τaos provides core infrastructure services and utilities essential for the broader TAO ecosystem. The subnet delivers foundational capabilities that other subnets and applications build upon, including cross-subnet communication, shared resources, and ecosystem coordination. Miners operate infrastructure while validators ensure service reliability.

RedTeam specializes in adversarial AI security testing, simulating attacks to identify vulnerabilities before malicious actors can exploit them. The subnet provides penetration testing automation, attack simulation, and security validation services. Miners develop offensive security tools while validators ensure tests are comprehensive and results accurate.

StreetVision by NATIX provides street-level computer vision for autonomous navigation, mapping, and urban analytics. The subnet processes dashcam and street imagery to create crowdsourced maps, detect road conditions, and support autonomous driving applications. Part of the NATIX ecosystem building decentralized physical infrastructure networks.

Brain subnet explores brain-inspired AI architectures and neuroscience-informed machine learning approaches. The subnet develops models mimicking biological neural networks, including spiking neural networks and neuromorphic computing. Miners implement novel architectures while validators benchmark against cognitive performance metrics.

OMEGA Labs is building the world's largest decentralized multimodal AGI dataset with over 1 million hours of footage and 30 million+ video clips covering 50+ scenarios and 15,000+ action phrases. Miners search YouTube and contribute video clips with ImageBind embeddings for video, audio, and captions. The Focus Videos marketplace enables users to earn TAO by recording task-completion videos purchased by miners.

BetterTherapy provides AI-powered mental health support and wellness coaching services. The subnet offers compassionate conversation, coping strategies, and emotional support tools while maintaining appropriate boundaries for professional care. Miners deliver supportive interactions while validators ensure responses meet safety guidelines.

Talisman AI enables personalized AI companion interactions with customizable character personalities and persistent memory. Users create unique AI personalities that remember conversation history and maintain consistent traits across interactions. The subnet powers engaging conversational experiences for entertainment, companionship, and creative roleplay applications.

FlameWire delivers high-speed data transfer and network communication infrastructure for low-latency applications within Bittensor. The subnet optimizes packet routing and bandwidth allocation to minimize communication delays between participants. Miners operate network nodes while validators benchmark throughput and ensure routing efficiency.

ReinforcedAI specializes in reinforcement learning, training agents through reward-based learning to master complex decision-making tasks. The subnet provides infrastructure for developing RL agents including environment simulation and training orchestration. Miners run training workloads while validators evaluate agent performance.

MASA's Subnet 42 powers decentralized data collection using Trusted Execution Environment (TEE) workers for secure data handling. Simple Docker-based deployment supports both miner and validator roles with automatic image pulls. Features include cookie refresh systems for Twitter authentication across Docker and Kubernetes environments, enabling reliable social media data access.

Omron serves as a dynamic hub, connecting users to specialized services within Bittensor subnets. Enabling access to diverse offerings like data, computational tasks, 
and complex analyses, while Omron ensures the origins of AI outputs are cryptographically verified through zero-knowledge proofs.

NI Compute powers a decentralized GPU marketplace where miners contribute resources and earn rewards while validators measure performance and allocate compute to clients. The subnet removes the 256 UID limit through executor spawning, enabling massive scaling of available computing power. Features include real-time metrics dashboard, DDoS protection, and integration with the Neural Internet cloud platform.

SN25 is a dedicated protein-folding subnet that leverages distributed compute, to reduce the cost per query and accelerate scientific discovery. By incentivizing miners to develop efficient protein-folding models, SN25 empowers researchers with powerful, low-cost tools. This is just the beginning—SN25 showcases how decentralized computing can tackle some of science’s toughest problems on the blockchain.

ReadyAI provides a low-cost structured data pipeline transforming raw data into tagged, semantic data for vector databases and AI applications. The fractal data mining approach enables miners to process diverse data sources while validators establish ground truth through full tagging. Scoring uses cosine distance between miner outputs and validator ground truth, leveraging LLMs that are now cheaper and more accurate than human annotators.

The Sportstensor subnet is designed to incentivize the discovery of competitive advantages such as 'edge' over closing market odds, enabling top miners within the network to establish machine-driven dominance across the sports prediction landscape.

RESI provides decentralized residential proxy infrastructure, giving access to verified residential IP addresses for web scraping, data collection, and geo-distributed research. Miners contribute residential connections while validators verify IP authenticity and connection quality. The network offers privacy-preserving access to residential IPs across diverse geographic locations.

Bitsota focuses on satellite and space data analysis, processing remote sensing imagery and orbital data for Earth observation applications. The subnet provides infrastructure for analyzing satellite imagery, tracking orbital objects, and deriving geospatial intelligence. Miners process space data while validators verify analysis accuracy.

Plaτform delivers foundational platform services enabling applications and services to build on top of Bittensor infrastructure. The subnet provides essential capabilities including hosting, resource management, and cross-service coordination. Miners operate platform infrastructure while validators ensure service reliability and performance.

VoidAI handles edge cases, null states, and exceptional conditions in AI systems with robust error recovery. The subnet specializes in graceful degradation when systems encounter unexpected inputs or failure modes. Miners develop resilient handling strategies while validators test against diverse edge case scenarios.

Tiger ALPHA specializes in alpha generation and market opportunity discovery, identifying high-potential investments before they become mainstream. The subnet uses AI to scan markets, analyze patterns, and surface asymmetric opportunities. Miners develop alpha signals while validators track prediction accuracy.

Internet of Intelligence (IoI) creates a globally connected network of AI systems sharing knowledge and capabilities. The subnet enables AI-to-AI communication, distributed learning, and collective intelligence emergence. Miners operate connected AI nodes while validators ensure network coherence and knowledge quality.

Dogelayer brings DOGE-inspired innovation and community-driven development to Bittensor with layer 2 scaling capabilities. The subnet combines meme culture's creativity with serious infrastructure development, creating an engaging environment for experimentation. Miners contribute to fun yet functional development while validators ensure quality.

Rich Kids of TAO provides premium services and exclusive features catering to high-value participants in the Bittensor ecosystem. The subnet offers enhanced capabilities, priority access, and specialized services for significant stakeholders. Miners deliver premium experiences while validators maintain service standards.

Oneoneone embraces simplicity and minimalism in AI operations, delivering streamlined, efficient services without unnecessary complexity. The subnet focuses on doing one thing exceptionally well with elegant, clean implementations. Miners optimize for simplicity while validators reward efficient, uncluttered solutions.

Minotaur tackles complex problem-solving inspired by the mythical labyrinth guardian, navigating intricate challenges that require strategic thinking. The subnet excels at problems with many interconnected factors and no obvious solutions. Miners develop navigation strategies while validators evaluate solution quality.

Taonado drives rapid transformation and accelerated development within the TAO ecosystem, enabling fast iteration and quick deployment of new capabilities. The subnet acts as a catalyst for ecosystem innovation, reducing time from concept to implementation. Miners deliver rapid development while validators ensure quality isn't sacrificed for speed.

Level 114 brings gamification and play-to-earn mechanics to Bittensor, creating engaging experiences where participation earns rewards. The subnet implements achievement systems, level progression, and competitive elements to make mining and validation fun. Miners earn through gameplay while validators manage fair competition.

SoulX provides digital identity and reputation infrastructure including soulbound credentials and proof of personhood. The subnet enables verifiable identity claims without exposing private information, building on-chain reputation through demonstrated actions. Miners process identity verification while validators maintain trust standards.

TaoLend operates as a decentralized lending protocol enabling TAO holders to lend and borrow against their holdings. The subnet creates efficient money markets with algorithmically determined interest rates based on supply and demand. Miners provide liquidity while validators ensure protocol security and fair rate calculation.

BrainPlay delivers cognitive training and brain games designed to improve mental fitness through engaging gameplay. The subnet provides scientifically-informed exercises targeting memory, attention, problem-solving, and other cognitive abilities. Miners create training content while validators evaluate effectiveness metrics.

HODL ETF provides ETF-style diversified investment exposure with long-term holding strategies for crypto portfolios. The subnet offers balanced, professionally-managed exposure across multiple assets without active trading requirements. Miners manage portfolio allocation while validators ensure strategy adherence and performance.

Akihabara is a decentralized visual intelligence network built for next-generation AI-native creators focused on anime and manga-style content. Unlike traditional image models, it specializes in sequence and narrative continuity where every panel, character, and scene can remember its own past. The subnet empowers artists, storytellers, and game studios to generate continuously evolving 2D manga-style worlds directly on-chain.

Affine is an incentivized reinforcement learning environment that rewards miners for making incremental improvements on reasoning tasks like program abduction and coding. The mechanism is designed to be sybil-proof, decoy-proof, copy-proof, and overfitting-proof, ensuring fair competition. Validators evaluate models deployed on Chutes (SN64) looking for the pareto frontier - the model that outcompetes all others across all environments.

Sundae Bar is a Bittensor subnet focused on treat and reward distribution mechanisms within the decentralized network ecosystem. The subnet creates innovative incentive structures for rewarding network participants based on contribution quality and consistency. Miners work to optimize reward allocation strategies while validators ensure fair and accurate distribution of incentives.

Bitrecs is a novel recommendation engine built on Bittensor providing real-time e-commerce product recommendations using large language models. The subnet addresses the 20T global e-commerce market by giving smaller retailers access to recommendation technology previously only available to giants like Amazon. Miners compete using LLM-based voting ensemble methods to produce quality recommendations that maximize merchant sales.

MANTIS is a Bittensor subnet that rewards miners for embeddings which help predict financial instrument targets with a 1-hour horizon. The system uses time-lock encryption to prevent miners from seeing future prices before submitting, ensuring fair competition. Validators collect encrypted embeddings, align them with price moves, and assign weights based on salience using XGBoost aggregation and permutation importance.

Swarm is a Bittensor subnet engineered to enable decentralized autonomous drone flight through reinforcement learning policies. Validators create synthetic map tasks and evaluate miner-supplied pre-trained RL policies inside a PyBullet physics simulator. The subnet aims to establish open-source flight algorithms as the go-to control intelligence for micro-drone navigation in research and industry.

ByteLeap is a distributed compute resource platform connecting GPU providers with the Bittensor network through a three-tier architecture of validators, miners, and workers. Miners aggregate worker resources and earn rewards through active compute leases and computational challenges including CPU/GPU matrix multiplication benchmarks. The scoring system combines lease revenue (70%) with challenge performance (30%) and availability metrics.

test short description