01 AI

What is 01.AI?

01.AI is a company that specializes in large models and agents, serving corporate and industry clients; it was founded by Kai-Fu Lee and is headquartered in Beijing. The company’s focus at present is not on ordinary chatbots, but rather on creating actionable decision-making and business management systems by combining models, agents, corporate data, and business processes.

Its product portfolio includes WanCe Enterprise AI Decision Hub, Boss AI, Investment Officer AI, Sales Champion AI, as well as the WanZhi Enterprise Large Model Platform designed for model deployment and fine-tuning. Zero One Everything has also introduced the Yi series of models, and it makes some of these models, codes, and technical materials available publicly on platforms such as GitHub.

The core product of Zero Ten Thousand Things

1. Wanci Enterprise’s AI Decision-Making Hub

The WanCe platform is used to link a company’s operational facts, business entities, decision-making logic, and action tasks. It aims to elevate AI from a tool for one-off queries to a decision-making hub that can identify problems, form judgments, monitor execution, and conduct continuous reviews.

2. Boss AI

Boss AI is designed for business owners and senior managers; it can be used for operational analysis, risk identification, and ensuring that actions are taken as needed. The system integrates internal company data and management principles to help management identify changes in business operations and track tasks related to improvement.

3. AI Investment Officer

The investment officer AI is designed for use in investment, M&A, corporate strategy, and board decision-making; it can assist with the analysis of business plans, due diligence materials, financial models, and industry research. It focuses on evaluating these materials, assessing risks, understanding valuation logic, and asking key questions, but it does not replace the investment committee in making decisions.

4. SalesKing AI

PinGuon AI is designed for B2B and B2G sales teams; it enables the analysis of the quality of business opportunities, customer needs, and the suitability of proposed solutions. Companies can use it to assist in allocating pre-sales resources, develop effective sales strategies, and keep track of the progress of key projects.

5. Magic Enterprise Large Model Platform

The Magic Platform covers aspects such as enterprise model deployment, application development, and model fine-tuning, and is suitable for technical teams that need to develop internal AI applications. It integrates with industry agents, data governance, and security mechanisms to help enterprise-level projects move from the prototype stage to production.

Main functions and technical capabilities

• Enterprise fact database: Integrates business systems, documents, metrics, project data, and publicly available external information.
• Business ontology modeling: Abstracting people, assets, processes, events, and decision rules into computable objects.
• Multi-agent collaboration: Allows different agents to handle tasks such as retrieval, analysis, risk assessment, finance, or business operations separately.
• Chain of evidence and review: Retain the basis for judgment, so that important recommendations can be examined against the facts and relevant materials.
• Action closure: Transform the analysis results into responsibilities, tasks, milestones, and follow-up actions.
• Model deployment: Choose between the cloud, a dedicated environment, or other delivery methods based on data, security, and computing power requirements.
• Model fine-tuning: Adjusting the model’s performance by using corporate datasets, industry knowledge, and task samples.
• FDE co-creation: Frontline deployment engineers enter the business environment to work together with clients to analyze requirements and test solutions.
• Yi series models: offer capabilities such as support for Chinese and English languages as well as handling long contexts, while also contributing to the development of an open model ecosystem.

Key industries and application scenarios

• Business management: Analyze revenue, costs, organizational, and operational metrics to identify issues that require management attention.
• Investment and M&A: Organizing project documentation, assessing risks, analyzing valuations, and assisting in the preparation of decision-making materials.
• Sales management: Assessing the quality of business opportunities, matching appropriate solutions, and improving the efficiency of pre-sales resources.
• Supply chain and logistics: Develop industry-specific intelligent agents focused on planning, order fulfillment, inventory management, and handling exceptions.
• Manufacturing and Energy: Handling production and operational tasks by combining equipment, processes, and specialized knowledge.
• Agriculture and retail: Connecting business data, industry expertise, and intelligent agents to specific operational scenarios.
• Sovereign AI: Develops AI infrastructure capabilities for national or regional-level data, models, computing power, and industrial applications.

How can enterprises make use of Zero One Everything?

1. Identify a business problem whose outcomes can be quantified, such as reducing the time required for due diligence or increasing the conversion rate of key business opportunities.
2. Identify the relevant personnel, systems, documents, data definitions, and existing decision-making processes.
3. Determine the combination of strategies, intelligence tools, and industry-specific agents in collaboration with the product or delivery team.
4. A prototype is developed using masked samples, and business experts review the facts, reasoning, and output structure.
5. Configure account permissions, data boundaries, operation logs, manual review processes, and exception handling mechanisms.
6. Run it in a small-scale real-world business environment to record the accuracy rate, processing time, adoption rate, and business outcomes.
7. Valid rules and cases that have been verified are turned into organizational knowledge, which is then gradually extended to more departments.

How to deploy and test the Yi model

1. First, determine whether a model needs to be made available, whether local deployment is required, whether platform APIs are needed, or whether enterprise-level delivery services are sufficient.
2. View the context, hardware requirements, quantized versions, and license of the target Yi model.
3. Use public examples to test Chinese comprehension, long documents, reasoning, and industry-specific tasks.
4. Conduct security testing, injection protection measures, content review, and performance evaluation prior to production deployment.
5. Estimate the total cost of GPUs, storage, operations and maintenance, as well as model invocations, based on concurrency and response time.

Prices and cooperation methods

The current official website of Zero One Ten Thousand Things focuses on business consultations and project collaborations; it does not list any fixed packages for services such as Wan Ce, Boss AI, Investment Officer AI, Sales Champion AI, or the Wan Zhi platform. The actual costs generally depend on factors such as data integration, the size of the models, the method of deployment, the number of users, industry-specific requirements, and implementation services.

Yi model and open-source status

The enterprise products and delivery platform related to ZeroOneTen thousand Things are not entirely open source; however, the developers have made available some of the models in the Yi series, examples of inference processes, methods for fine-tuning, as well as various resources related to this ecosystem. The Yi project offered versions with different sizes – 6B, 9B, 34B – as well as versions designed for chat, basic tasks, visual processing, or handling long contexts.

The “open model” concept does not mean that all versions have exactly the same licensing terms. Before using a model for commercial purposes, redistributing it, modifying it, or providing online services, companies should refer to the licenses and usage terms available in the respective model repository.

Which customers are suitable?

• Medium and large enterprises that wish to pursue an AI transformation by focusing on management decisions and operational results.
• Investment firms that need to establish a chain of evidence for investment research, due diligence, and decision-making.
• Sales teams in companies that have complex B2B or B2G sales processes.
• Organizations that need a combination of industry agents, corporate knowledge, and proprietary data.
• Customers who need integrated services for model deployment, fine-tuning, application development, and security management.
• Developers and algorithm teams interested in researching or deploying the Yi series of open models.

Product advantages

• The focus is clear: addressing high-value business issues related to operations, investment, and sales.
• It not only provides models but also covers consulting, data, agents, deployment, and co-creation of business solutions.
• It emphasizes facts, evidence, risks, and a closed loop of actions, making it suitable for situations requiring serious decision-making.
• It has experience in developing Yi-series models as well as a certain open model ecosystem.
• It covers industries such as manufacturing, energy, agriculture, supply chains, investment, education, and retail.

Usage restrictions and precautions

• There is no fixed, public price for corporate projects, and the initial communication phase, as well as the pilot and implementation stages, can be quite lengthy.
• The effectiveness of the system depends heavily on the quality of the company’s data, its business rules, and the level of involvement of experts.
• Decision-making AI can only assist in making judgments; investment, management, and legal responsibilities remain the duty of the company’s employees.
• When accessing financial, customer, and transaction data, it is necessary to strictly configure access permissions and apply data masking.
• The historical evaluations of the open-source Yi model cannot directly reflect the actual performance of current enterprise products.
• There may be version differences between the older API platforms, earlier models, and the current range of enterprise products.

Support methods

• Corporate Web Platform and Management Backend
• Integration of enterprise data with business systems
• Integration of model APIs and agent capabilities
• Deployment in the cloud, in a proprietary environment, or according to the specifications set for each project
• Local or on-premises deployment of the Yi open model
• Consulting, FDE co-creation, and delivery of industry solutions

Basic information

Recommendation score

4.2 / 5. ZeroOneEverything is more suitable for enterprise customers who have clear business objectives, a solid data foundation, and a dedicated implementation team, rather than ordinary users who are simply looking for free chatbots.

Frequently Asked Questions

Is Lingyi Wanshu a chatbot?

That’s not all. The company is currently focusing on enterprise decision-making platforms, industry-specific intelligent agents, large-model platforms, and services for enterprise AI transformation.

What are the main products of Zero One Ten Thousand Things?

It mainly includes the Wance Strategy Platform, Boss AI, Investment Officer AI, Sales Champion AI, as well as the Wanzhi Enterprise Large Model Platform, and it also offers industry-specific solutions and consulting services.

Is there a free version of Zero One Everything?

The company’s products do not offer any public, unified free subscription plans. Some models from the Yi series can be downloaded and deployed under corresponding licenses, but this incurs costs related to computing power, development, and maintenance.

Can the Yi model be used for commercial purposes?

It is necessary to check the license for the specific version of the model. The terms vary depending on the model, weights, and distribution channel; therefore, the license of one repository cannot be used as a reference for all Yi products.

Is Zero One Ten Thousand suitable for small teams?

If a small team only needs basic Q&A functions, an enterprise-level solution might be overkill. Teams that have specific industry use cases, proprietary data, and particular deployment requirements would benefit from further consultation.