Edge AI Solutions For Enterprise
Edge AI Solutions For Enterprise
企业端侧私域 AI 方案供应商 Enterprise edge and private AI solution provider
具芯科技 juxin.ai 专注于高性价比端侧 AI 硬件、推理框架、基础优化模型和内置智能体，先替客户解决最复杂的本地部署、多模态推理与知识库底座，再由生态伙伴和行业用户补齐场景 skills 与应用，让 AI 真正进入千行百业。 Juxin.ai focuses on cost-effective edge AI hardware, inference frameworks, optimized foundation models, and built-in agents. We solve the hardest deployment, multimodal inference, and knowledge base problems first, then let ecosystem partners and industry teams extend the final scenario skills and applications.
我们是谁
Who We Are
面向客户提供端侧 AI 整体软硬件产品方案
We provide complete edge AI hardware and software product solutions
我们交付什么
What We Deliver
硬件、推理框架、优化模型与内置基础智能体一体交付
Hardware, inference stack, optimized models, and built-in agents delivered together
我们怎么扩展
How We Scale
复杂底座由我们解决，场景 skills 由生态与行业用户持续完善
We solve the hard shared foundation, then ecosystem and industry teams extend the scenario skills
Solutions
Solutions
先把底座做好 再让场景落地 Build the foundation first then let scenarios scale
我们的价值不是只卖一台盒子，而是把端侧硬件、推理框架、基础模型、内置智能体和 skills 生态组织成一套可以交付、可以扩展、可以进入真实业务现场的端侧 AI 方案。 Our value is not just selling a box. We package edge hardware, inference frameworks, foundation models, built-in agents, and a skills ecosystem into one deliverable system that can expand into real business scenarios.
端侧软硬一体 Integrated Edge Stack
从高性价比端侧硬件，到推理框架和基础模型，我们把客户最难搭起来的底座先做成标准产品。 From cost-effective edge hardware to inference frameworks and base models, we standardize the foundation customers struggle to assemble.
内置基础智能体 Built-in Core Agents
文档理解、知识检索、实时感知和基础助手能力先内置好，让客户拿到设备就能开始真实业务验证。 Document understanding, retrieval, realtime sensing, and baseline assistants come built in so customers can start real validation immediately.
Skills 生态扩展 Skills Ecosystem
通用复杂度由我们承担，细分行业用户和生态伙伴专注补齐场景 skills 与应用，形成面向千行百业的扩展路径。 We handle the hard shared complexity, while ecosystem and vertical teams focus on scenario skills and applications for industry expansion.
Panoramic Architecture
Panoramic Architecture
Juxin.ai 做底座 客户与生态做场景 Juxin.ai builds the base customers and ecosystem build scenarios
底部三层是具芯负责的共享复杂底座，包括硬件、推理框架、基础模型和内置智能体；最上层则由客户和生态伙伴围绕销售、客服、巡检、制造、零售、办公等场景持续扩展 skills 与应用。 The lower shared foundation is built by Juxin.ai across hardware, runtime, optimized models, and built-in agents. The top layer is expanded by customers and ecosystem partners into sales, support, inspection, manufacturing, retail, office, and many more scenario skills.
客户与生态把场景做深做广
Customers and ecosystem deepen and expand the scenarios
销售
Sales
客服
Service
巡检
Inspection
制造
Manufacturing
零售
Retail
办公
Office
知识
Knowledge
培训
Training
用户系统与 Skills
User System and Skills
场景助手、行业应用、生态扩展件
Scenario agents, apps, and ecosystem extensions
OpenClaw 与私有 RAG
OpenClaw and Private RAG
内置智能体、工具调用、知识检索
Built-in agents, tool orchestration, and retrieval
模型与推理框架
Models and Inference Runtime
优化模型、多模态运行时、推理底座
Optimized models, multimodal runtime, and inference base
MoFang 01 硬件与算力
MoFang 01 Hardware and Compute
端侧部署、局域网、NAS 与设备交付
Edge deployment, LAN, NAS, and device delivery
Products
Products
MoFang 01 是我们的首款端侧 AI 产品 MoFang 01 is our first edge AI product
它不是一台只会聊天的设备，而是一台企业本地多模态智能中枢。能看文档、读图片、理解表格、处理语音，并结合本地知识库给出更贴近业务的回答。 It is not just a chat device. It is a local multimodal enterprise intelligence hub that can read documents, images, tables, and voice while grounding responses in local knowledge.
MoFang 01 Flagship
MoFang 01 Flagship
企业本地多模态智能中枢 A local multimodal intelligence hub for enterprise
MoFang 01 强调本地部署、开机即用、数据留在本地。标准版承接文档智能与知识检索，Pro 面向实时感知与现场交互，Pro Max 负责更多助手、更多插件与更完整的企业级私有化部署。 MoFang 01 is built around local deployment, ready-to-use onboarding, and local data control. Standard focuses on document AI and retrieval, Pro on realtime sensing and interaction, and Pro Max on richer assistants, plugins, and enterprise-scale deployment.
本地部署
Local Deployment
数据留本地
Data Stays Local
Intel Arc Pro B60
30+ tokens/s
Key Specs
Key Specs
一台设备 让企业本地 AI 真正跑起来 One device to get local enterprise AI running
Intel i7
企业级 CPU 底座
Enterprise CPU base
32GB
内存
Memory
Arc Pro B60
24GB GDDR6 显存
24GB GDDR6 VRAM
30+ tok/s
3 路模型并发
3-way model concurrency
这些参数的意义不是堆配置，而是让大模型、本地知识库和常见办公资料处理真正可以在客户现场稳定运行。 These specs are not for show. They make local models, knowledge bases, and common enterprise document workflows run reliably on site.
Typical Scenarios
Typical Scenarios
一台设备覆盖多种业务动作 One device supports multiple business actions
快查
Fast Search
合同、制度、报价、历史资料秒级定位。
Find contracts, policies, quotes, and prior material in seconds.
会写
Draft Faster
销售文案、报价说明和方案摘要快速成稿。
Draft sales copy, quote notes, and solution summaries quickly.
能答
Answer Better
客服知识库和常见问答更一致，新人也能更快上手。
Keep support knowledge and FAQs consistent so new staff ramp faster.
会看
See More
图片、扫描件、表格和语音一起理解，更接近真实办公资料。
Read images, scans, tables, and voice together for real office material handling.
SKU 01
MoFang 01
面向文档智能、知识检索、售前演示与轻量私有化部署，是客户最快启动端侧 AI 的入门版本。 Built for document AI, retrieval, presales demos, and lightweight private deployment as the fastest entry point into edge AI.
- 标准化设备形态，适合快速 PoC 与样板客户部署 Standardized device format for fast PoCs and lighthouse deployment
- 优先承接文档助手、本地知识库与日常资料快查 Best for document assistant, local knowledge base, and fast information lookup
SKU 02
MoFang 01 Pro
面向实时感知、语音交互、巡检值守与现场多模态协作，更适合需要设备参与业务现场的客户。 Built for realtime sensing, voice interaction, inspection, and live multimodal workflows where the device participates directly in field operations.
- 增强外设接入与现场语音、视频、图像理解能力 Stronger peripherals plus voice, video, and image understanding on site
- 适合门店、展厅、产线、客服和巡检等一线场景 Fits retail, showroom, production, support, and inspection scenarios
SKU 03
MoFang 01 Pro Max
面向企业级完整部署、多助手协同、更多插件与更复杂场景并发，是复杂项目和长期落地的旗舰版本。 Built for full enterprise deployment, multi-agent coordination, and richer plugin concurrency as the flagship option for complex long-term projects.
- 为复杂项目预留模型、skills 与行业应用扩展空间 Leaves room for model, skills, and vertical application expansion
- 适合长期驻场、跨部门共享与更深度的私有化交付 Fits long-running deployments, cross-team sharing, and deeper private delivery
Customer Value
Customer Value
MoFang 01 带来更快成交 更稳服务 更强沉淀 MoFang 01 brings faster wins steadier service and reusable value
它把企业资料、经验和流程沉淀为可复用的智能能力，让销售、客服、运营和管理岗位更快进入统一的智能化工作入口。 It turns enterprise materials, know-how, and workflows into reusable intelligence so sales, service, operations, and management teams can work through one AI entry point.
成交更快 Faster Deals
销售找资料、写方案、做报价更快，客户响应更及时，缩短跟进与成交周期。 Sales teams find material, draft proposals, and produce quotes faster, improving response speed and shortening the path to close.
服务更稳 Steadier Service
客服和售后基于统一知识库答得更一致、更专业，新员工也能更快上手。 Support teams answer more consistently from one knowledge base, and new staff ramp up faster.
知识更值钱 Knowledge Becomes Asset
企业资料、经验和流程不会只留在个人脑子里，而会逐步沉淀成公司的智能资产与业务护城河。 Enterprise materials, experience, and processes stop living only in individuals and become reusable intelligence assets for the company.
硬件与算力 Hardware & Compute
MoFang 01、Intel GPU、局域网与 NAS 构成交付基础，让客户先用合理成本把端侧 AI 真正跑起来。 MoFang 01, Intel GPU, LAN, and NAS form the delivery base so customers can get edge AI running at a sensible cost.
模型层 Model Layer
基础优化好的文档、多模态和 Embedding 模型先内置好，让客户不必从零调模型和调推理参数。 Optimized document, multimodal, and embedding models come ready so customers do not need to tune models and runtime from scratch.
OpenClaw 与 RAG OpenClaw & RAG
内置基础智能体、工具调用和私有知识检索，负责把模型能力直接变成可工作的业务流。 Built-in agents, tool orchestration, and private retrieval turn model capability into working business flows.
用户系统 User System
Skills 市场、场景助手和行业应用负责把通用能力延伸到具体业务现场，走入真正的千行百业。 The skills marketplace, scenario agents, and vertical applications extend the shared platform into real industry use cases.
News
News
Juxin.ai 最新新闻 Latest News From Juxin.ai
新闻区不只是更新列表，它负责让客户看到我们正在把产品、交付和生态同步往前推进。 News is more than an update list. It shows customers that product, delivery, and ecosystem are moving forward together.
MoFang 01 开始形成完整产品表达 MoFang 01 is taking shape as a complete product story
从标准版、Pro 到 Pro Max，MoFang 01 不再只是硬件命名，而是对应不同客户阶段、不同业务深度和不同交付复杂度的一套产品线。 From Standard to Pro and Pro Max, MoFang 01 is no longer just hardware naming. It is a structured product line mapped to different customer stages and delivery depth.
查看产品结构 View Product Structure一期继续坚持轻底座 Phase one keeps a light base
统一认证、Docker 白名单代理、ClawHub 同步与 OneAPI 开源网关仍然是第一阶段交付主线。 Unified auth, Docker whitelist proxy, ClawHub sync, and an open-source OneAPI gateway remain the primary delivery line in phase one.
基础智能体与 skills 市场开始成形 Core agents and the skills marketplace are taking shape
具芯负责通用复杂底座，行业团队与生态伙伴则能围绕不同场景继续做深应用层。 Juxin handles the shared complex foundation so vertical teams and ecosystem partners can deepen the application layer.
客户交流口径正在统一 Customer-facing messaging is being unified
后续官网、销售演示、PoC 方案和部署文档将使用同一套产品结构和价值表达。 The website, sales demos, PoC plans, and delivery documents will share one product structure and value narrative.
Contact
Contact
开始试用 方案咨询或部署沟通 Start a trial solution consultation or deployment discussion
如果你已经准备开始测试，可以直接进入平台；如果你还在评估场景，也可以先从试用申请、方案沟通和部署咨询开始。 If you are ready to test, go straight into the platform. If you are still evaluating the scenario, start with a trial request, solution discussion, or deployment consultation.
Business Routes
Business Routes
联系我们 Contact Us
试用申请
Trial
trial@juxin.ai
方案咨询
Solution
solution@juxin.ai
部署支持
Support
support@juxin.ai
联系电话
Phone
020-83706770
注册地址
Registered Address
广州市黄埔区光谱中路11号云升科学园3栋5层05+06单位
Units 05+06, 5F, Building 3, Yunsheng Science Park, No. 11 Guangpu Middle Road, Huangpu District, Guangzhou
Next Step
Next Step
进入统一入口 或先预约方案沟通 Enter the platform or book a solution session
如果你已经准备测试平台，直接进入统一登录。如果你还在评估场景，先通过方案沟通确认设备、模型和交付节奏。 If you are ready to test the platform, enter unified login now. If you are still evaluating scenarios, start with a solution conversation to align devices, models, and delivery pace.