BitTorrent launched BTTInferGrid to build a decentralized AI inference computing power foundation, poised to drive a comprehensive leap in BTT value.

On June 17th, BitTorrent, the world's leading decentralized file transfer ecosystem, announced the launch of its core AI strategic product, BTTInferGrid, building a decentralized computing network for AI inference scenarios.

BTTInferGrid is a major AI product strategically upgraded by BitTorrent based on its mature decentralized storage service, BTFS. It consolidates BitTorrent's deep technical expertise accumulated over years in core areas such as P2P network protocol design, global distributed node governance, and large-scale resource scheduling. This gives the platform inherent advantages for large-scale application and commercial implementation from its inception. The official launch of this product not only marks the beginning of BitTorrent's foray into the decentralized AI infrastructure赛道, but also officially opens a new chapter for distributed computing power empowering the development of the AI industry.

Relying on its cryptoeconomic incentive system and distributed consensus mechanism, BTTInferGrid seamlessly connects globally idle GPU computing resources with the diverse inference needs of AI developers. It provides next-generation AI applications with open, verifiable, pay-as-you-go efficient inference services, while also allowing idle GPU holders to easily monetize their assets, creating a win-win situation for both computing power supply and demand.

From a technical infrastructure perspective, BTTInferGrid reconstructs the traditional centralized computing power supply system through distributed resource aggregation and intelligent scheduling mechanisms, granting AI infrastructure greater resource elasticity and resilience against risks. From an industry landscape viewpoint, it frees computing power from its scarce and monopolistic nature, transforming it into a freely circulating digital production asset. This allows every GPU holder to participate in value creation and revenue distribution, fostering a new industry paradigm of universal access to shared computing power and efficient circulation.

BitTorrent Launches BTTInferGrid: Building the Decentralized AI Inference Computing Base 

"Computing power, algorithms, and data" are the three core elements of AI development, and the strategic value of computing power has reached unprecedented heights in 2026. The "computing power shortage" is no longer just a long-term industry warning but has evolved into the primary bottleneck hindering the advancement of AI.

Looking at the global market, rental costs for high-end NVIDIA GPUs continue to rise, and hardware supply remains chronically tight. Leading AI companies like OpenAI and Anthropic frequently experience server outages due to insufficient computing power reserves. Even tech giants and top academic institutions are struggling to secure adequate computing resources. SpaceX, which recently listed on Nasdaq, admitted in its IPO prospectus that the computing power demands of its AI systems have significantly outpaced current market supply, even considering reclaiming computing resources previously lent to Anthropic for self-preservation. Meanwhile, Microsoft's cloud platform Azure was reportedly forced to urgently seek computing power from competitor Amazon AWS to cope with the massive demand gap caused by the surge in code submissions on GitHub in the AI era. At the same time, AI labs at top universities like Stanford and MIT have had to suspend multiple large model training projects due to computing power shortages, even causing delays in many graduate students' thesis defenses.

It is against this backdrop of escalating global supply-demand contradictions for computing power that BTTInferGrid was created. It aims to build a decentralized AI inference computing power network (DePIN). By aggregating scattered, idle GPU computing resources globally in a decentralized manner, it precisely matches the business needs of AI developers, breaks down the barriers and monopolies formed by traditional centralized computing power service providers, maximizes the utilization of global idle hardware resources, and establishes a new generation of affordable, open, and shared computing infrastructure. This fully unleashes the potential of global idle hardware, ensuring every unit of computing power is utilized and maximizing its value.

To ensure the efficient implementation of the entire operating system, BTTInferGrid adopts a modular, layered architecture design, building a three-tier collaborative system: "Application Layer - Computation Layer - Settlement Layer":

• Application Layer: Serving as the service entry point for developers, the Application Layer provides a user-friendly deployment environment, supporting the rapid implementation of various AI-native applications, such as AI chatbots, intelligent agents, and other diverse scenarios.
• Computation Layer: As the core computing hub of the entire ecosystem, the Computation Layer bears the critical responsibilities of performing AI model inference computations, handling real-time request responses, and managing task scheduling.
• Settlement Layer: The Settlement Layer is responsible for the automated operation of the entire economic system, covering the entire process including computing power staking, task settlement, contribution reward distribution, and malicious node penalties. This layer executes on-chain transactions in a trustless manner, ensuring fair and transparent value exchange between computing power suppliers and demanders without the need for intermediaries, providing a solid foundation of economic trust for the entire network.

The three layers collaborate efficiently through standardized interfaces: the Application Layer initiates inference requests, the Computation Layer schedules computing resources to execute them, and the Settlement Layer automatically distributes incentives based on the execution results. The three layers support each other in a closed loop, collectively forming a high-performance, highly trustworthy, and sustainably developing decentralized AI inference infrastructure. 

Based on this three-tier architecture, BTTInferGrid offers multiple advantages, including distributed node autonomy, demand-driven permissionless access, and end-to-end verifiable trustworthiness, establishing an efficient, stable, open, and barrier-free distributed computing environment.

From a network architecture perspective, BTTInferGrid adopts a globally distributed node deployment strategy. All nodes are collectively owned and maintained by the community, with no single data center or operating entity controlling the network core. This inherently decentralized design completely avoids the common single points of failure and operational interruption risks associated with traditional centralized platforms. It grants the network strong censorship resistance and 7x24 uninterrupted service resilience, providing a highly available operational base for various AI inference tasks.

Regarding the access and scheduling rules for computing resources, BTTInferGrid implements a permissionless, open mechanism: any GPU device meeting the performance standards can freely join the network without requiring approval from a centralized authority. Furthermore, the overall computing power supply is entirely driven by real business demand. Incentive calculations are based on the node's actual computing power utilization and comprehensive service performance, complemented by a dynamic supply adjustment mechanism that flexibly allocates resource scale based on real-time network load. This mechanism not only improves the turnover efficiency of computing resources but also ensures that computing power providers receive stable long-term income commensurate with their contributions.

At the trust mechanism level, BTTInferGrid embeds trust logic throughout the entire business process. Leveraging a robust cryptoeconomic system, the network can automatically perform operations such as computing power scheduling, task assignment, and revenue settlement. Every AI inference computation task is fully traceable, and computation results support on-chain cross-verification. Through the design of its underlying mechanisms, the network fundamentally prevents violations like false computing power reporting or data tampering, ensuring the authenticity and integrity of all computational tasks, allowing demanders to use the service with confidence and suppliers to participate securely.

In summary, the distributed node architecture provides the computing network with autonomy and high stability; the demand-driven, permissionless access model ensures efficient resource turnover and long-term economic sustainability; and the end-to-end verifiable trust system safeguards the ecosystem's security baseline. The deep integration of these three core features makes BTTInferGrid not just a technologically advanced distributed computing network, but a long-term stable, highly trustworthy, and future-oriented decentralized AI infrastructure. 

BTT Poised to Become the Core Value Token of the Decentralized AI Computing Network, Potentially Broadening Ecosystem Application Boundaries 

As BitTorrent's native value token, with the official launch and continuous expansion of the BTTInferGrid ecosystem, BTT's strategic positioning may undergo a critical upgrade. Its application scenarios could expand radically from traditional distributed transmission and storage tracks to the entire AI computing infrastructure value chain, continuously broadening the ecosystem's value boundaries.

In the past, BTT was the circulation medium for BitTorrent, the world's leading decentralized file transfer network. Now, leveraging the new AI computing network BTTInferGrid, it is poised to evolve into the core token for scheduling the decentralized AI computing network, assuming the dual functions of value transfer and ecosystem governance.

The cryptoeconomic incentive mechanism of BTTInferGrid is the underlying engine driving the network. It connects off-chain idle GPU computing power with the inference needs of AI developers. Through token incentives, it automates task scheduling, result verification, and revenue settlement, ensuring transparent supply-demand matching and governance. 

Within the BTTInferGrid system, the continuous operation of the ecosystem mainly relies on the synergistic participation and division of labor among three core roles: miners, users (AI developers), and validators. Together, they build an autonomously operating decentralized computing network:

• Miners (Computing Power Suppliers): Contribute idle GPU resources, undertake and execute AI inference tasks, and earn rewards based on actual workload, task completion quality, and dynamic performance scores.
• AI Developers (Computing Power Demanders): Access the global distributed computing pool via a unified, standardized API, significantly reducing the cost of calling upon computing resources.
• Validators (Network Guardians): Audit and randomly challenge the computational performance of miner nodes, identifying cheating, low-quality computing power, and other anomalous behaviors. They earn rewards for maintaining network security and service quality.

These three types of participants form a complete, symbiotic, and mutually constraining closed loop based on a decentralized consensus mechanism, jointly driving the continuous evolution and virtuous cycle of the BTTInferGrid ecosystem. The core link connecting the rights and interests of all parties and driving the ecosystem's healthy operation is the customized cryptoeconomic incentive system built specifically for BTTInferGrid.

Through the circulation of tokens, this system achieves precise quantification and fair distribution of computing power value, converting actions like supplying computing power, executing tasks, and auditing results into clear, quantifiable incentive signals. Miners earn token rewards for contributing idle GPUs and completing inference tasks with high quality. Validators earn income by maintaining network security. AI developers pay fees based on actual computing power consumption. The interests of the three parties achieve dynamic balance through the circulation of the token economy, thereby creating a sustainable value closed loop.

Within this framework, BTT is expected to become the unified native incentive and settlement base token within the BTTInferGrid ecosystem, running through the core aspects of the entire computing power ecosystem. It will comprehensively cover the processes of paying for AI computing resource usage, incentivizing contributions, and dynamically allocating resources, ultimately building a closed-loop economic system where "computing power contributors receive rewards, computing power users pay conveniently, and ecosystem participants share value."

Specifically, BTT tokens can play multiple core roles within the BTTInferGrid network: As a payment medium, AI developers use BTT (or its equivalent token) to pay for inference service fees, achieving "on-demand procurement, pay-as-you-go." As an incentive tool, miners receive token rewards based on verified actual computational contributions, while validators earn income for providing audit and challenge services, continuously attracting global idle resources to the network. As a staking asset, validators must stake tokens to participate in scoring and verification; computing nodes also need to stake a certain number of tokens to qualify for undertaking tasks. Any misconduct triggers a stake slashing mechanism, effectively ensuring network security and fairness from an economic standpoint.

From this perspective, BTT is not only poised to be the value vehicle matching computing power supply and demand but also the fundamental core driving force underpinning the efficient, fair, and long-term operation of the entire decentralized AI computing economy. On one hand, token incentives continuously attract more idle GPU resources to join the network, expanding computing power supply. On the other hand, the supporting staking and slashing mechanisms ensure the stability and reliability of inference services. Furthermore, all settlement, reward, and penalty logic is automatically executed by smart contracts, effectively resolving the common pain points of information opacity and high trust costs prevalent in centralized computing power platforms.

As the BTTInferGrid ecosystem develops and thrives, BTT is expected to become the universal value anchor connecting distributed computing power and AI application demands, ushering in a new paradigm for the decentralized AI economy.

BTTInferGrid Reconstructs Global Computing Power Distribution, BitTorrent Opens a New Chapter in the Decentralized AI Track

 Against the backdrop of intensifying global AI computing power supply-demand conflicts and growing monopolization by centralized providers, BTTInferGrid reconstructs the computing power supply model through distributed technology. It efficiently aggregates fragmented, idle GPU resources worldwide, building an open and shared computing infrastructure. This allows AI developers to access elastic computing power with zero barriers while enabling every unit of idle computing power globally to realize its inherent value. Simultaneously, leveraging its innovative cryptoeconomic incentives and collaborative governance mechanisms, it opens up a value circulation closed loop between the computing power supply side and the demand side, forming an ecological cycle where both sides mutually promote and positively reinforce each other.

For Miners (Computing Power Suppliers), BTTInferGrid acts as a "value converter," transforming idle computing power into a source of sustainable income. Any idle GPU meeting the basic performance threshold can join the network permissionlessly, contributing computing power to earn rewards.

Distinguishing itself from the crude models of traditional distributed computing platforms that simply distribute rewards based on "raw hardware computing power," BTTInferGrid employs a multi-dimensional scoring, weighted incentive model. The network comprehensively evaluates core metrics such as the node's actual effective workload, task response latency, service stability, and result accuracy. It then dynamically calculates and distributes corresponding rewards. This mechanism fundamentally breaks the monopoly of "large computing power dominating rewards," allowing small and medium-sized miners who provide high-quality, reliable services to also earn excess returns, institutionally guaranteeing the overall service quality of the network. Additionally, miners who participate in the network early will enjoy ecosystem benefits like exclusive reward multipliers, gaining a first-mover advantage.

For AI Developers, BTTInferGrid provides AI inference computing power services that are openly accessible, verifiable, and flexibly priced on a pay-as-you-go basis. It offers a computing solution completely different from traditional cloud vendors, effectively addressing prevalent industry pain points such as "expensive computing power, poor elasticity, and difficult trust," significantly lowering the barrier to entry for deploying AI applications.

Firstly, BTTInferGrid offers elastic computing power scheduling, dynamically allocating resources based on AI inference load. Developers no longer need to pre-purchase hardware or sign long-term contracts, completely freeing themselves from resource lock-in by centralized cloud vendors and achieving true on-demand usage and flexible scaling. Secondly, it adopts a decentralized market-based pricing model with precise token-based billing, eliminating the high premiums charged by centralized platforms, significantly reducing inference costs, and bringing computing power expenditures back to reasonable levels. Most critically, BTTInferGrid builds a decentralized multi-validator auditing network. Through multiple mechanisms like random challenges, cross-verification, and stake slashing, it technically prevents computing power fraud and result tampering, ensuring every inference computation is authentic, traceable, and its results are verifiable. These multiple advantages complement each other, making BTTInferGrid not only a cost-effective channel for acquiring computing power but also a decentralized AI inference infrastructure trusted by developers.

In terms of product development, BTTInferGrid has formulated clear, actionable short-term, medium-term, and long-term development plans to steadily advance the iterative upgrade and ecological expansion of the decentralized AI computing network:

Short-term Goal (2026) Focuses on network launch and basic service implementation. While gradually increasing the number of online GPU nodes, it will complete core node go-live and inference service validation, add support for mainstream open-source models like DeepSeek and Qwen, and launch API services for developers and enterprise clients;

Medium-term Goal (2027) Aims at ecosystem closure and capability boundary expansion. Based on the stable operation of inference services, it will comprehensively enhance network performance and ecosystem richness, achieving an upgrade from a single inference service to a comprehensive computing platform (e.g., model fine-tuning, cross-chain resource access), building a complete developer toolchain and ecosystem support system;

Long-term Goal (2028 and beyond) Aims to become an AI-native infrastructure, building a collaborative network integrating computation, storage, and smart contracts. It will provide foundational support for AI agents and automated applications, ultimately becoming the preferred decentralized inference layer for global open-source AI applications, providing elastic, affordable, and trustworthy computing power support for large-scale, high-concurrency next-generation AI application scenarios.

In terms of ecosystem development, BTTInferGrid has already completed native adaptation for several top-tier open-source large models in the industry. This includes mainstream models like Alibaba Cloud's Tongyi Qianwen Qwen3.6 27B, Qwen2.5 7B Instruct, and Meta's Llama 3.1 8B Instruct, covering diverse business scenarios like general conversation, code generation, and content creation. Developers do not need to deploy and debug models themselves; they can flexibly call upon them on demand through standardized API interfaces, further lowering the barrier to entry and significantly shortening the development and launch cycle for AI applications.