News related to
cloud

At the heart of the showcase was Phancy’s “Token Factory”, which demonstrated a complete closed-loop from technological innovation and engineering implementation to industrial empowerment. The exhibit received widespread acclaim from industry leaders, government and enterprise representatives, international guests, and major media outlets. A number of high-profile guests visited the booth, sparking in-depth discussions on innovation opportunities and development trends in the AI 2.0 era.

Building the “Token Factory”: Full-Stack Capabilities Driving Scalable Industrial Adoption

At WAIC 2026, Phancy centered its presentation on its full-stack AI cloud service platform, demonstrating the “Token Factory” and turning the vision of scalable, reusable and deployable AI productivity into tangible industrial capabilities.

At the infrastructure and foundational layer, Phancy’s HAMi (virtual heterogeneous GPU management and orchestration platform) and ModelHub XC (opensource model adaption and optimization platform) effectively addresses key industry pain points — including localized deployment of large models, efficient computing resource scheduling and large‑scale implementation. These solutions provide a solid foundation for industrialized AI production and strengthen the Token Factory’s core advantages in efficiency, stability, and scalability. Both technologies attracted significant attention at the event, drawing government officials, industry experts, and enterprise clients for demonstrations, consultations, and partnership discussions.

Core Production and Intelligent Applications

At the core production and application layer, Phancy highlighted its AGI platform Phanthy Pantheon and the self-evolving AI modeling tool PhanthyModel, both of which significantly lower the barriers to building advanced models. On the delivery side, the company introduced its cross-platform embodied agent framework PhanthyMotus, the AI coding agent PhanthyCode, and the AI-powered film production platform PhanthyMovie. These solutions enable seamless integration of AI capabilities into real-world operations across a wide range of industries.

From computing orchestration and model development to intelligent agent deployment, Phancy continues to connect computing resources, AI models, developers, and industry ecosystems, driving the stable and efficient transformation of AI technologies into practical, scalable industrial productivity.

Internationally Recognized as a Trusted Global Industrial AI Solution

In addition to its technology showcase, Phancy received significant international recognition at WAIC 2026. At the Global Industrial AI Cooperation Forum, in the presence of high-level guests including the UN Deputy Secretary-General, Brazil’s Minister of Public Service Management and Innovation, and academicians from the Chinese Academy of Sciences, Phancy was selected as a “Global Flagship Case in Industrial AI” and named among the first batch of “Trusted Global Industrial AI Solutions.”

This prestigious honor validates Phancy’s technological reliability, compliance and real‑world implementation capabilities in industrial AI, and marks strong international endorsement of the company’s AI engineering and commercialization strengths.

Building an Open Industry Ecosystem and Entering a New Era of AI 2.0

The successful conclusion of WAIC 2026 was not only a celebration of technological achievements but also a powerful platform for deep industry collaboration. As AI moves into the era of large-scale application, Phancy will continue to pursue technological innovation and open collaboration, further strengthening its full-stack AI cloud service system.

Looking ahead, Phancy will work closely with global developers, industry partners, and clients to bring the “Token Factory” into more real-world scenarios, making AI a true driver of industrial upgrading and high‑quality productivity.

Hashtag: #PhancyGroup

The issuer is solely responsible for the content of this announcement.

The launch combines a unified, action-oriented world model with Ambient Capture Engine 2.0, three commercial solutions and a new industry benchmark, connecting high-density data with real-world robotic operations.

SHANGHAI, CHINA – Media OutReach Newswire – 19 July 2026 – ACE ROBOTICS today unveiled Kairos 3.1, its latest action-oriented world model, alongside Ambient Capture Engine 2.0 and three commercial solutions spanning instant retail, hospitality and outdoor service scenarios.

The announcements were made at a forum ACE ROBOTICS organized during the 2026 World Artificial Intelligence Conference, convened around the theme of advancing physical AI from “understanding” to “execution.” The event brought together economists, technology executives and embodied AI researchers to examine how general-purpose world models can move from laboratory research to industrial deployment.

The forum also marked the launch of PHYSICAL IQ, a unified benchmark for embodied physical intelligence. It was jointly initiated by the Shanghai Artificial Intelligence Association, the Shenzhen Loop Area Institute and the East China branch of the China Academy of Information and Communications Technology, with participation from more than 20 universities and industry partners.

Thomas J. Sargent, Nobel laureate in Economic Sciences, discussed the limitations of current intelligent systems in rare and previously unseen situations.

A world model built for the physical world

“In the digital world, a model error may result in a flawed image or paragraph. In the physical world, an incorrect action can have real consequences,” said Wang Xiaogang, Chairman of ACE ROBOTICS. “Kairos 3.1 is built around a first-principles approach to embodied world models, helping robots act more reliably in complex and uncertain environments, and accelerating the arrival of physical AI’s Kairos moment.”

Kairos 3.1 is designed as a natively unified model that integrates generative, physical and cognitive intelligence within a single architecture, rather than combining separately developed capabilities.

Built on a hybrid Transformer architecture with a shared mixed-attention mechanism, it brings visual observations, language instructions, force and tactile signals, and policy trajectories into a unified latent space.

This supports an “understand, reason, execute and reflect” loop. The model can break down long-horizon tasks, simulate physical cause and effect across multiple scenarios, rank candidate actions, execute the selected strategy and evaluate the outcome for further adjustment.

At the core of its spatial understanding is ACE-BRAIN-0.5. ACE ROBOTICS said it has achieved state-of-the-art results across 12 public evaluations covering spatial understanding, navigation, manipulation and task-progress assessment, among publicly reported models as of July 2026.

In a household laundry scenario, for example, a robot can identify spatial relationships between objects, divide a task into more than a dozen steps and verify each stage in real time. If a failure occurs, it can identify the affected step and restart from that point rather than repeating the entire task.

For physical generation and reasoning, Kairos-HomeWorld supports whole-home scene generation and object interaction. It is built on 300,000 residential floor plans, 5,000 simulated home environments and 8,700 3D assets covering six categories of physical properties, designed to reflect common residential layouts in China.

The model can simulate multiple action trajectories in parallel and rank them based on predicted success and execution cost before sending instructions to a physical robot.

In internal testing, the Kairos 3.1 8B model achieved an inference latency of 125 milliseconds on the NVIDIA Jetson Thor platform at BF16 precision. Its in-house KairosRT computing engine supports real-time, on-device inference.

Kairos 3.1 also incorporates self-reflective iteration. When an action fails, the robot can evaluate the result and adjust its strategy. In one test, for example, it changed from a three-finger to a four-finger grasp after an unsuccessful attempt and subsequently completed the task.

From high-density data to continuous learning

ACE ROBOTICS also introduced what it calls the “Information-Density Law” for embodied models: the value of data is determined not only by its volume, but by whether it contains information capable of changing the outcome of an agent’s actions.

ACE ROBOTICS classifies embodied data across five information-density levels, from L1 to L5. L1 and L2 data support foundational pre-training; L3 and L4 data remain focused on known tasks; and at L5, data incorporates three-dimensional force and tactile signals, failure-recovery trajectories and variables from open environments. The company says these higher-density forms of interaction data support stronger generalization, reflection and continuous learning.

Built on this principle, Ambient Capture Engine 2.0 is a human-centric system for capturing, processing and reusing high-density physical-interaction data.

ACE Ego Kit is a lightweight wireless wearable comprising head, hand and chest components. It includes the ACE Sense Glove, which offers sensitivity of 0.01 newtons and a joint-angle error of less than two degrees. The system can synchronize data from more than 20 heterogeneous sensors with a timing error of less than one millisecond.

ACE Data Engine is an automated data-production platform for continuous, high-precision annotation of long-horizon tasks. It incorporates ACE-ViDiHand, a generative 4D hand-motion-capture framework designed to track movements despite occlusion and rapid motion, which ACE ROBOTICS reports led across three public benchmarks (ARCTIC, HOT3D and HOI4D) as of July 2026. The company recorded a frame-level accuracy of 0.997 and up to a 4.8-fold improvement in motion smoothness.

ACE Ego Matrix standardizes embodied data across four dimensions: spatial coordinates, embodiment structure, action timing and data quality. It is designed to align data collected by different operators, devices and robot platforms for reuse across systems. ACE ROBOTICS has open-sourced the framework, which it says placed first on the RoboCase and RoboTwin leaderboards.

ACE ROBOTICS also announced the open release of ACE-Data-0, an L5 household-interaction dataset featuring complex physical tasks and high-precision annotations.

Three solutions moving into commercial operation

Building on its data infrastructure and Kairos world-model capabilities, ACE ROBOTICS introduced three standardized industry solutions designed for deployment within existing commercial workflows.

Xiaoman, its integrated fulfillment solution for instant retail, is paired with the new W1 fulfillment robot. The W1 has a robot-to-payload weight ratio of less than 2:1, force-control precision within one newton and a minimum required aisle width of 75 centimeters, making it suitable for high-density shelving environments. It is able to grasp both flexible packaging and rigid goods.

The solution connects robot hardware, world-model capabilities, motion control, product information, warehouse locations and order management. It supports lightweight deployment within days across flash-fulfillment warehouses, small-format stores and integrated warehouse-store setups. It has been deployed with customers including Sense MartGo, Kuaikeda and PetroChina convenience stores, using real order and shelf data to improve fulfillment performance. ACE ROBOTICS plans to deploy the solution across 1,000 retail locations over the next year and expand to approximately 10,000 stores within two years.

Xiaoxin addresses hotel-laundry operations, where ACE ROBOTICS estimates that approximately 80% of demand occurs at night.

The solution enables robots to handle collection, loading, unloading, washing, sorting and folding, with ironing capabilities planned for future versions. It is designed to reduce night-shift staffing pressure and improve operational consistency across non-standard hotel laundry environments.

Xiaotu is designed for autonomous operation in complex outdoor environments. Using a “one brain, many bodies” architecture, it applies a general-purpose intelligent brain (Agentic AI) across multiple quadruped robot platforms.

The system supports indoor and outdoor navigation, dynamic obstacle avoidance, autonomous charging, remote dispatch and multi-robot coordination. It has been deployed in visitor-guidance and interactive experience services at cultural-tourism events, including the Begonia Flower Festival in Tianjin.

Building an open embodied AI ecosystem

ACE ROBOTICS also announced a series of partnerships spanning infrastructure, computing and commercial deployment.

The company entered into a strategic collaboration with the Caohejing Development Zone to develop an embodied AI innovation platform.

In retail and fulfillment, ACE ROBOTICS announced collaborations with PetroChina, Kuaikeda and SenseTime Shanhui covering autonomous warehouse stores, front warehouses and unmanned retail environments.

To support the computing requirements of world-model development, ACE ROBOTICS launched the World Model Cloud Ecosystem initiative with Baidu AI Cloud, Alibaba Cloud, Huawei Cloud, Tencent Cloud and SenseCore AI Cloud.

By connecting high-density data, unified world models, robot platforms, commercial applications and shared evaluation standards, ACE ROBOTICS aims to support the transition of embodied AI from individual technical demonstrations to sustained operations in real-world environments.

Hashtag: #ACEROBOTICS

https://www.linkedin.com/company/acerobotics/posts/?feedView=all&viewAsMember=true%EF%BD%9C
https://x.com/ace_robotics

The issuer is solely responsible for the content of this announcement.

About ACE ROBOTICS – Equipping robots with intelligent “brains” and engaging “souls”

ACE ROBOTICS is a pioneering robotics company dedicated to advancing the field of embodied intelligence. Through breakthrough technological innovations and deep insights into embodied intelligence scenarios, we aim to empower robots with the ability to autonomously understand and explore the physical world, thereby accelerating their commercial implementation.

The company pioneered the ACE R&D paradigm and built a vision-based “environmental data engine, real-world cognition, embodied interaction generalization” technology chain. Using full spatiotemporal and multi-perspective environmental capture as its engine, along with Kairos 3.0 – China’s first open-source and commercially applicable world model – plus the Embodied Foundation Model as its technical backbone, ACE ROBOTICS addresses core industry challenges such as data scarcity, common sense gaps, poor generalization, and limited versatility. Simultaneously, the company unveiled its flagship A1 Embodied Super Brain Module, accelerating the large-scale commercial deployment of embodied intelligence across diverse scenarios.

ACE ROBOTICS is both a technology pioneer and an ecosystem builder. Through strategic cooperation with top hardware manufacturers, cloud service providers, and vertical scenario partners, we have broken through the “model-hardware-scenario” industrial deadlock, providing standardized and customized solutions that are driving the development of China’s embodied intelligence industry.

Link: Guizhou in July

In midsummer July, Malaysian travel operators and cultural tourism influencers visited Guizhou, China, a renowned summer retreat, as if stepping into another season. Known as the “Green Karst Kingdom” and a “Natural Air Conditioner,” Guizhou has a mountainous area coverage of 92.5% and an average summer temperature of just 23°C. Beyond its cool climate, Guizhou offers a hidden-gem experience where spectacular landscapes meet diverse cultures, leaving every visitor amazed.

Link: https://youtu.be/nL7ayiYqB-k?si=nfasDLnhaBIFgkqB

Spectacular Landscapes at Every Turn

From the lush peak forests of Wanfenglin to the magnificent Huangguoshu Waterfall; from the emerald waters of Xiaoqikong to the cloud-covered summit of Mount Fanjing; and to the world-record Huajiang Grand Canyon Bridge spanning a deep gorge, every landscape in Guizhou is worth a stop. A Malaysian travel influencer remarked, “Before departure, I had only heard that Guizhou was a summer retreat. After visiting, I realized the surprises go far beyond 23°C.”

Cultural Experiences Full of Surprises

Visitors can enter distinctive villages and try intangible cultural heritage crafts such as Miao embroidery and batik; gather around a long-table banquet to taste local specialties such as sour soup fish and spicy chicken; explore cool hidden landscapes within a World Natural Heritage site with children through waterfall play and paddleboarding; or venture deep into ancient caves for an underground adventure. From heritage crafts to nature exploration, Guizhou offers family-friendly experiences for everyone to enjoy.

A Spontaneous Trip Made Easy

This summer, reserve a ticket for Mysterious Guizhou, and begin your “summer escape plan.”

Explore more of Mysterious Guizhou: www.guizhoutravel.com

Hashtag: #Guizhou

The issuer is solely responsible for the content of this announcement.

Dr. Benson Wu, CEO and Co-Founder of CyCraft, stated: “We believe, being listed as a Sample Providerin the Gartner report demonstrates that the global market now views AI security as the next battlefield for AI commercialization. From day one, CyCraft has developed proprietary models, simultaneously deploying AI reasoning capabilities across three core pillars: cybersecurity defense, AI governance, and unmanned vehicle reconnaissance. Technical leadership is our starting point, not the destination. Our goal is to translate this leadership into rapid growth driven by strategic deployment and disciplined execution. Launched this year, our AI Red Teaming and AI Guardrail solutions have already been productized and commercially shipped, seeing active deployment in sectors like semiconductors and sovereign AI.”

AI Reasoning Models Drive Up Operating Costs, Restructuring Profit Logic

As AI transitions from “training models” to “reasoning models,” the industry officially moves from the research and development phase into commercial application. Reasoning models possess logical deduction, self-reflection, and complex problem-solving capabilities. However, these advanced capabilities have caused AI operational costs to skyrocket, forcing the tech industry to redefine its expectations of AI as a tool for massive cost reduction and profit maximization.

CyCraft’s XCockpit Threat Exposure Management Platform features built-in AI reasoning models that triage thousands of alerts in real time, automatically providing mitigation recommendations and effectively eliminating alert fatigue. From overall security posture scoring to specific actionable tickets for individual alerts, XCockpit transforms defensive benefits—which were historically difficult to define—into concrete, quantifiable technical metrics. This embodies CyCraft’s pragmatic philosophy: the value of defense must be visible, and it must be measurable.

Standing at the Frontier of AI Security, CyCraft Advances Commercial Implementation with Profound R&D Momentum

Deeply integrating AI with cybersecurity, CyCraft has launched XecGuard and XecART—two major AI security products deployable across both cloud and on-premises environments, both featured on the OWASP AI Security Map. These solutions directly address the core pain points of Agentic AI, LLM, and GenAI applications. XecART utilizes Agentic Red Teaming to run multi-round stress tests on models, identifying risks and generating compliance reports that adhere to international standards. Meanwhile, XecGuard acts as a real-time Guardrail for LLMs and AI Agents, intercepting anomalous inputs and outputs. One solves “how to stress-test AI,” while the other solves “how to protect AI,” providing the fundamental AI governance for enterprises.

CyCraft’s research achievements are equally outstanding. The CyCraft research team recently presented at the International Conference on Machine Learning (ICML), one of the world’s top three prestigious academic conferences for AI and machine learning. The research focused on Retrieval-Augmented Generation (RAG) Data Poisoning. Utilizing their self-developed “Eyes-on-Me” attack framework, the team proved that attacks can simultaneously breach both the retrieval and generation phases. The end-to-end average attack success rate surged from 21.9% to 57.8%. This breakthrough exposed a long-neglected vulnerable attack surface existing between “attention focus” and “model output.”

In 2025, the CyCraft research team was also selected for NeurIPS, one of the most iconic academic sanctuaries in the AI field. The team proposed an innovative “unified supervision framework” that embeds verifiable checkers into synthesized instruction-conflict instances, enabling self-supervised learning without requiring oracle labels or reasoning traces. This framework not only applies to different models but also significantly enhances both safety robustness and instruction adherence.

CyCraft believes that as AI moves from R&D to commercial deployment, the barrier for enterprises is no longer “whether they can use AI,” but “whether they can use AI responsibly.” Once Agentic AI enters financial transactions, government services, or critical infrastructure, every autonomous decision must leave a traceable path, accept auditing, and allow for real-time human intervention when necessary. This is true AI governance. It is not a compliance document, but a complete control chain stretching from pre-launch to runtime, and onto post-incident traceability.

CyCraft deconstructs this control chain into three actionable phases: XecART stress-tests model risks before deployment; XecGuard safeguards inputs and outputs during runtime; XCockpit quantifies the overall risk posture into traceable metrics. Interlocking these three components ensures that every step AI takes is visible, manageable, and auditable. CyCraft’s objective is clear: make AI behavior measurable before its value can be priced. Ultimate technical leadership must be converted into commercial resilience that is governable, auditable, and open to investor scrutiny.

The technological revolution of AI Reasoning Models and Agentic AI is an undeniable reality. The ultimate challenge lies in whether enterprises can implement “controllable, auditable, and highly transparent” AI governance within real-world commercial environments. From being recognized in international research reports to taking the stage at top-tier academic conferences, CyCraft continues to prove that from visionary R&D to product commercialization, Taiwan’s AI cybersecurity prowess helps herald the future of a secure, trusted, and prosperous AI era.

Reports
Gartner, Emerging Tech: AI Vendor Race: Reasoning Models Reshape Providers’ Approaches to AI Pricing, 5 June 2026.

Disclaimer
Gartner does not endorse any company, vendor, product or service depicted in its publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner publications consist of the opinions of Gartner’s business and technology insights organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this publication, including any warranties of merchantability or fitness for a particular purpose. GARTNER is a trademark of Gartner, Inc. and its affiliates.

Hashtag: #CyCraftTechnology

The issuer is solely responsible for the content of this announcement.

About CyCraft