BTC $63,067.40 -1.58%
ETH $1,833.61 -2.71%
BNB $563.39 -2.28%
XRP $1.08 -1.86%
SOL $74.70 -1.74%
TRX $0.3220 -0.40%
DOGE $0.0718 -1.76%
ADA $0.1605 -0.98%
BCH $222.39 +0.77%
LINK $8.18 -2.85%
HYPE $60.25 -8.49%
AAVE $90.54 -4.30%
SUI $0.7347 -0.81%
XLM $0.1838 -2.52%
ZEC $533.78 -2.46%
BTC $63,067.40 -1.58%
ETH $1,833.61 -2.71%
BNB $563.39 -2.28%
XRP $1.08 -1.86%
SOL $74.70 -1.74%
TRX $0.3220 -0.40%
DOGE $0.0718 -1.76%
ADA $0.1605 -0.98%
BCH $222.39 +0.77%
LINK $8.18 -2.85%
HYPE $60.25 -8.49%
AAVE $90.54 -4.30%
SUI $0.7347 -0.81%
XLM $0.1838 -2.52%
ZEC $533.78 -2.46%

decent

All
Article
Flash

X-Agent officially releases Litepaper V1, building a decentralized AI agent economy

The AI Agent no-code operating platform X-Agent, based on the Web3 social network, officially releases its core "Litepaper V1" to the world. The white paper details how it reconstructs the decentralized AI agent ecosystem and commercialization path through no-code deployment and secure sandbox technology.The core technologies and solutions are as follows: 1) Speak to Build (no-code deployment): Allows users to create and deploy professional-grade AI agents to Telegram, WhatsApp, or the web in minutes using pure natural language. 2) SRE Cryptographic Sandbox (physically isolated security): A unique physically isolated operating environment that perfectly protects user private keys and sensitive credentials while enabling agents to have enterprise-level autonomous asset changes and on-chain collaboration capabilities.Market performance and latest data: X-Agent has surpassed 1,000,000+ cumulative global users, and the AI agents have autonomously completed over 1,100,000 actual business tasks, consuming/destroying over 8.4 billion model tokens. The project has previously raised $1.8 million in funding and has a highly cohesive localized community in Japan and South Korea. The native token $XAGT serves as the network's settlement hard currency, directly used for SRE computing Gas fee settlement, transaction commission payment, and staking endorsement. The team and investors have a strict 12-month Cliff lock-up. The community quota is completely non-inflationary and can only be unlocked through actual sandbox task consumption algorithms, ensuring that each token is supported by real business demand. According to the latest roadmap, X-Agent will open a new no-code public portal in Q3 2026 and launch a decentralized Agent Store in Q1 2027.

The decentralized lending protocol Goldfinch, supported by a16z, announced that it will gradually shut down

According to Cryptopolitan, the decentralized lending protocol Goldfinch, supported by a16z, announced it will gradually shut down. Last Friday, an anonymous investor, Edward Morra, publicly accused the protocol of mismanagement, leading to over $50 million in user fund losses, stating that borrower defaults and failed loan restructurings made it nearly impossible for depositors to recover their funds. Just a day after the post was published, the project announced it would enter a gradual shutdown phase.The protocol's native token GFI fell from a peak of $32.94 in January 2022 to below $0.07, a decline of 99.8%, with its market capitalization dropping from over $390 million to less than $6 million. Goldfinch was founded in 2021 by former Coinbase employees, aiming to connect crypto capital with credit enterprises overlooked by traditional banks, with a16z leading its $25 million financing in January 2022. Problems began to emerge months after the funding: the Kenyan motorcycle financing company Tugende Kenya defaulted, two underlying positions in the $2 billion loan from the U.S. credit fund Stratos nearly went to zero, and the Singapore borrower Lend East could only repay 58% of the principal. As the loan portfolio deteriorated, the protocol turned to institutional credit funds but ultimately could not turn the situation around.

first_img Bittensor co-founders release decentralized roadmap, aiming to complete it within a year and a half

Bittensor co-founder const posted on the X platform, detailing the current state of decentralization of the project, future roadmap, and goals. Bittensor has not yet achieved decentralization at the economic incentive layer, which is still led by the core team, including const himself, two engineers, and a core group of contributors. The project has been live for over 5 years, with no pre-mining, and has 128 subnet teams and over 20 core validator teams, achieving decentralization at the ownership distribution level. The team chose to rapidly iterate at the cost of "maintaining centralization" rather than slowly advancing "democratic" decision-making.Regarding future update plans, Bittensor will promote validators to re-enter the competitive mechanism while opening liquidity pools for two-way investment to symmetrize the market and prevent on-chain signals from being manipulated. Additionally, a belief mechanism will be introduced to grant voting rights to Alpha token holders. In the coming weeks, updates will also be made to TaoFlow and its derivatives, further fine-tuning the issuance distribution algorithm to optimize the distribution method of inflation. Const expects to complete the construction of core mechanisms within the next year and a half, at which point the three pillars of incentive alignment, value optimization, and true ownership will operate in synergy, ultimately achieving complete decentralization by abandoning centralized control.

Hyperbridge restarts the cross-chain interoperability protocol and launches the OFT adapter, completing the decentralized architecture upgrade

Hyperbridge Cross-Chain Interoperability Protocol Hyperbridge announced the completion of a comprehensive architecture reboot, re-launching after completing security audits, bug bounty incentives, and system reconstruction, and officially transforming into a "hyperstructure." The protocol was suspended after the security incident on April 13, during which it completed a joint audit with organizations such as SRLabs and paid over $150,000 in bounties to security researchers.The team stated that this upgrade removed the original centralized management keys, transitioning to a fully permissionless network of validators and provers, achieving full-stack decentralized operation.This reboot also introduced the "Hyper Fungible Token (HFT)" standard, making each cross-chain asset an independent application layer structure, with issuers autonomously controlling cross-chain behavior rules, including pause mechanisms and throttling strategies. Hyperbridge also released the OFT (Omnichain Fungible Token) adapter, which is compatible with existing cross-chain protocols like LayerZero, allowing assets to be migrated to a zero-knowledge proof-based transport layer by simply modifying configuration parameters, without the need to redeploy contracts.In addition, the protocol's business model has shifted from a pay-per-use model to a subscription model, allowing cross-chain applications to pay a stablecoin fee of $50 to $1,000 per month for bandwidth services. The official statement indicated that this upgrade marks Hyperbridge's transition from an early cross-chain bridge project to a fully decentralized infrastructure protocol, aiming to provide a unified interoperability layer for a multi-chain ecosystem without the need for trusted intermediaries.

DGrid AI released the latest research paper PoQ-Judge, completing the closed loop of decentralized LLM quality assessment with a multi-architecture evaluation framework

The decentralized AI infrastructure network DGrid AI today released its latest research paper "PoQ-Judge," proposing a multi-architecture quality assessment framework that does not require reference answers. This means that in real deployment environments, there are often no standard answers for comparison, yet the protocol can still reliably score the quality of model responses and allocate incentives accordingly. This is a key piece that has long been missing in DGrid's decentralized LLM inference quality assessment system.PoQ (Proof of Quality) is a consensus mechanism independently developed by DGrid, designed to prevent model providers from deploying low-quality models, fabricating data, or hiding computational costs at the protocol level, thereby ensuring service quality and pricing transparency. The DGrid team has been continuously working on PoQ and has published four research papers to date. The newly released PoQ-Judge has trained three assessment models covering different quality and cost scenarios, achieving a correlation of up to 0.747 with human scoring on the retention test set, significantly outperforming all previous reference answer-based evaluators, while reducing assessment costs by over 72% through cascading evaluation and online weight calibration.With the implementation of PoQ-Judge, the entire process from quality assessment → scoring → incentive allocation has completely eliminated reliance on reference answers, thus establishing a closed loop for the quality of decentralized LLM inference.DGrid AI is a decentralized AI intelligent network dedicated to building an open, transparent, and community-driven AI infrastructure. Focusing on model invocation and application experience, DGrid has launched several core products: the AI Gateway that aggregates mainstream large models globally, the one-click deployment platform for AI agents DClaw, the anonymous model competition platform AI Arena, and the intelligent model recommendation assistant Dori, providing one-stop services for developers and users. It is reported that DGrid AI's revenue has surpassed 20 million dollars in six months.

DGrid AI has officially partnered with MiniMax to expand the decentralized model ecosystem

The decentralized AI infrastructure network DGrid AI announced an official partnership with MiniMax. Both parties will provide low-cost, high-availability access to MiniMax models for global developers and agents through the DGrid decentralized AI Gateway, jointly promoting its adoption in a broader ecosystem.DGrid is a decentralized AI infrastructure that aggregates mainstream AI models from around the world. It currently connects over 200 models, including Claude, GPT, and Gemini, enabling one-stop access through a unified API. The platform is aimed at developers and AI agent builders, offering high-availability, low-latency model access services, and is committed to lowering the entry barriers and usage costs for AI applications.MiniMax is a global leader in general artificial intelligence across all modalities and is one of the few AI companies capable of self-developing models for text, voice, video, music, and other modalities. This official collaboration with MiniMax represents another significant expansion for DGrid on the model supply side, marking a new step in connecting mainstream model vendors and building a decentralized AI ecosystem. With the integration of more high-quality models, DGrid is becoming a unified entry point for AI developers to access diverse model capabilities, providing more flexible and open infrastructure support for the next generation of intelligent applications.As part of the initial collaboration, MiniMax's flagship model M3 has been launched on DGrid, and DGrid Premium users can enjoy a 55% discount directly.
app_icon
ChainCatcher Building the Web3 world with innovations.