STARWARCLOUD STARWAR.AI
XINGZHAN TECHNOLOGY · ABOUT US

Explore the edge of intelligence, create endless possibilities.

Full-stack ownership is the only real control.

Meet XingzhanExplore technology & outcomes
AI ECOSYSTEMDIGITAL WORLD
星战科技 AI 生态数字宇宙

Positioning & brand promise

WHO WE ARE & WHY WE EXIST

Hangzhou Xingzhan Technology Co., Ltd.

Answering who we are and why we exist

An AI-native company. From the lowest layer of compute to digital employees that get work done, everything is built in-house.

AI has moved from 'can talk' to 'can work', yet four things still hold enterprises back: compute is hard to buy, models are hard to govern, data cannot leave the network, and employees struggle to adopt it. Xingzhan turns these four into one complete in-house chain.

AI should not be a technology an enterprise buys — it should be a capability that grows inside it.

PARENT COMPANY
Hangzhou Xinglin Technology Co., Ltd.
Hangzhou Xinglin Technology Co., Ltd. is the parent company of Hangzhou Xingzhan Technology Co., Ltd.
A leading provider of new liquid-cooled intelligent computing center solutions and integrated AI operations. Using immersion liquid-cooling, it customizes racks, deploys ultra-high-density servers, develops its own cloud platform and connects liquid-cooled data centers with high-density computing — delivering integrated hardware, software and compute-storage services.
DATA CENTER
Liquid-cooled data center footprint
We already operate liquid-cooled data centers in Hangzhou, Wuhu, Shanghai, Jinhua, Dingxi and other cities. In the future we plan low-carbon liquid-cooled data centers across eight regions: Beijing-Tianjin-Hebei, the Greater Bay Area, the Yangtze River Delta, Chengdu-Chongqing, Inner Mongolia, Guizhou, Gansu and Ningxia.
CONTACT & REGISTRY
Contact & registration
Address
9F, Building 3, Xixi Yintai City, Xihu District, Hangzhou
Phone
189-5810-2708
Email
18958102708@189.cn
Website
XingZhan.Cloud (https://xingzhan.cloud)
ICP filing
浙ICP备2025202852号-1

Why Xingzhan

Full-stack in-house
Compute, models, agents and scenarios are all owned by us. No assembly, no outsourcing — every layer can be adapted to the customer’s business.
Secure & controllable
Data, models and links are independently controlled. Core knowledge never leaves the intranet; permissions, auditing and encryption are defined by the customer. Security is not outsourced to a third party.
A composable product matrix
Eight products share one stack. Take one entry point or combine the whole set.
Private deployment (cost-controlled, validated delivery)
Software-hardware in one box and ready out of the box, or deployed into the customer’s existing data center.
99.9%
Service availability
Externally committed, subject to the signed service agreement
20+
GPU specifications
Heterogeneous management across vendors and chip architectures
390+
Mainstream models
Connected through the OpenAPI unified gateway
7×24
Technical support
Ticketing and emergency fault response

Technology philosophy & innovation approach

FROM POSSIBILITY TO VALUE

Technology beyond possibility, intelligence that truly creates value.

Our view of technology and the values behind it

The value of AI lies not only in stronger models, but in the effective connection between compute, tools, knowledge and business scenarios.

ExploreConnectCreate
01

Continuous exploration

We follow the development and application of AI technology so new capabilities keep entering real scenarios.

02

Connecting capabilities

We explore the collaboration between compute, models, tools, knowledge and applications to reduce the complexity of bringing technology into use.

03

Innovation that lands

Starting from real needs, we turn technology exploration into usable, iterable product experiences.

Core capability architecture

CORE CAPABILITY ARCHITECTURE

From foundational capability to intelligent applications.

Explaining how the AI ecosystem comes together

An AI ecosystem needs more than model capability: it needs coordination between compute resources, platform tools and application scenarios.

Xingzhan continues to explore AI infrastructure, platform capabilities and intelligent applications, connecting technical capabilities across scenarios and gradually turning them into usable, iterable product value.

These three layers describe Xingzhan’s own technology layout and do not mean every product shares one underlying system. The eight products share one stack and can be used individually or combined as needed.

Provides compute resources and runtime support for AI workloads.

  • Elastic, heterogeneous distributed compute management
  • Containerized training-inference integration
  • Multi-tenant isolation, metering and billing
GPUCloud · compute operations platform

Connects models, tools, knowledge and agent capabilities.

  • 390+ models behind one gateway with optimal routing
  • Multi-agent collaboration and role orchestration
  • Private deployment and knowledge retention
OpenAPI · token factory and relay platform

Creates product value for different users and business scenarios.

  • Enterprise work and organizational collaboration
  • University AI training and certification
  • Marketing growth and interactive experiences
View the full product ecosystem

Selected work & outcomes

SELECTED WORK & OUTCOMES

Where technology lands, value becomes visible.

Proving capability with real evidence

Real innovation is not only about the technology itself, but about how it solves problems, improves experiences and keeps creating value in real scenarios.

From technical practice to product application, we focus on how each exploration becomes verifiable outcomes.

CASE STUDY 01 · TECHNOLOGY IN PRACTICE

From technical capability to real-world application.

GPUCloud upgrades “data center management” to “compute operations”: it manages compute resources across vendors and chip architectures through four technical traits — cluster, distributed, heterogeneous and elastic — supports integrated training and inference, and underpins external operations with multi-tenant isolation, metering and full-link monitoring.

  • ClusterA single node failure does not affect the overall service; overcommit and priority preemption are supported
  • DistributedFailed nodes are automatically replaced and state is resynchronized without restarting the job from scratch
  • ElasticOnline workloads and offline training run in off-peak shifts; preemptible instances lower overall TCO
Learn about GPUCloud
CASE STUDY 02 · PLATFORM & SYSTEM CAPABILITY

One gateway to govern every model.

OpenAPI uses a unified gateway to hide vendor API differences, aggregates 390+ mainstream models, routes requests by real-time latency and response speed, and provides transparent billing and quota management.

View OpenAPI
CASE STUDY 03 · APPLICATION & SCENARIO INNOVATION

Put AI inside the enterprise’s own boundary.

WorkMesh is designed for on-premises deployment. It brings together capabilities employees can actually use and the organization’s accumulated knowledge assets, addressing core data leakage, compliance blind spots without auditing or DLP, and inverted value.

View WorkMesh

The items above describe the engineering practice and platform capabilities of released products; they are not customer success stories. Public customer cases still require written authorization, so no unauthorized customer names or outcome figures are shown here.

Product ecosystem & business footprint

PRODUCT ECOSYSTEM

Bringing AI capability into more real-world scenarios.

Showing how technology lands in business

From compute infrastructure to intelligent platforms, from enterprise work to creative experiences, Xingzhan keeps exploring how AI applies across different fields.

Xingzhan Technology
AI INNOVATION ECOSYSTEM

News & insights

NEWS & INSIGHTS

The latest progress, all in one place.

Company updates and industry insight

Recent updates and industry observations published by Xingzhan, covering product capability, industry practice and company news.

AiEdu-Lab showcase wall at the Digital Trade Expo

Industry

Digital trade competition wraps up: AiEdu-Lab brings AI teaching into the classroom

On September 27, the national final of the 5th National College Digital Trade Comprehensive Ability Competition concluded at the Digital Trade Expo. Xingzhan presented its in-house AiEdu-Lab platform on the partner showcase wall.

Illustration for OPC one-person companies and DiWorker

Deep dive

Xingzhan redefines OPC: the one-person company and the DiWorker open collaboration platform

The word “company” is quietly changing its meaning — it used to mean offices, people and headcount; in the future it may mean one thing only: can the work get done?

Actor He Shengming visiting the Xingzhan office

Industry

Actor He Shengming visits Xingzhan Technology

On the afternoon of August 18, 2026, actor He Shengming visited Xingzhan Technology, walking the office floor and learning about the team’s working life.

Strategic cooperation signing ceremony between Xingzhan and Blue AI

Industry

Xingzhan Technology and Anhui Blue AI announce strategic cooperation

Hangzhou Xingzhan Technology Co., Ltd. and Anhui Blue AI Technology Co., Ltd. held a strategic cooperation signing ceremony, focusing on AI education practice and university talent development.

Vision & next step

VISION & NEXT STEP

The edge of intelligence keeps extending through creation.

Closing the brand story and inviting the next step

AI is still evolving, and each new technical capability opens up new possibilities.

Xingzhan will keep exploring the connections between compute, platforms and intelligent applications, responding to real needs with technology and working with partners to turn more ideas into reality.

Explore our productsContact Xingzhan