A dynamic charging and resource consumption display method based on AI agent role differentiation
Patent Information
- Application Number
- CN202610914051.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-24
- Publication Date
- 2026-10-27
AI Technical Summary
[0002]现有的AI付费服务主要采用订阅制、按token计费或按对话次数计费三种模式，存在以下不足：计费方式与AI实际提供的价值脱节——查一次天气和做一次深度行业分析消耗资源不同但价格相同；资源消耗展示冰冷不直观，缺乏差异化定价策略；免费额度策略单一
[0007]有益效果：价值匹配——计费与AI实际提供的价值对齐，用户为价值付费而非为字符付费；体验友好——虚拟资源概念替代冰冷的金额数字，降低用户付费心理门槛；分层引导——免费额度引导用户完成首次体验转化为付费用户；品牌契合——虚拟资源名称与平台品牌风格一致。
Abstract
Description
Technical Field
[0001] This invention relates to the field of paid artificial intelligence services, specifically to a method and system for differentiated billing based on agent role, tool call type, and user behavior in an AI agent platform, as well as for displaying resource consumption using a virtualization-friendly concept. Background Technology
[0002] Existing paid AI services primarily employ three models: subscription-based, token-based, or per-conversation-based. These models suffer from several shortcomings: the billing method is disconnected from the actual value provided by the AI—checking the weather and performing in-depth industry analysis consume different resources but cost the same; resource consumption displays are impersonal and lack differentiated pricing strategies; and free quota strategies are simplistic. Therefore, a method is needed that can differentiate billing based on agent role and tool type, and display resource consumption in a user-friendly way.
[0003] A search revealed that existing technologies include a general token-based billing model, where all model calls are charged uniformly based on the number of tokens, without distinguishing between agent roles and tool types; and a subscription-based membership model with a fixed monthly fee, lacking flexibility. Existing billing models do not cover a complete solution of "differentiated pricing based on AI agent roles, plus virtual resource concept demonstrations and tiered free quotas." This application is novel and inventive. Summary of the Invention
[0004] This invention aims to provide a differentiated billing method and system for AI agents. By tiered pricing based on agent role and tool call type, displaying resource consumption using a virtualization concept, and implementing a tiered free quota strategy, it achieves transparent, user-friendly, and value-matched billing.
[0005] A dynamic billing method based on AI agent role differentiation includes the following steps: Step S1 – Billing Strategy Configuration: The system configures a billing level for each AI agent role and each tool type. The billing level is determined based on a comprehensive consideration of computing power cost, value positioning, and usage frequency. Step S2 – Virtual Resource Conversion: The system converts the monetary value of each billing level into a unified virtual resource unit. The virtual resource unit uses a friendly name consistent with the AI platform's brand style, including but not limited to energy value, bamboo leaf, power bar, and honey drop count. Step S3 – Billing Triggering and Display: When a user initiates an operation that consumes resources, the system determines whether it is within the free quota. If the free quota is exceeded, the corresponding amount of virtual resources is deducted according to the billing strategy, and the user is informed in a contextually embedded manner in the reply content without pop-ups or interrupting the conversation. If virtual resources are insufficient, the system prompts for replenishment in a friendly manner. Step S4 – Free Quota Management: The system configures a tiered free quota for each user, allocated according to three dimensions: agent role, time period, and operation type. Step S5 – Resource Consumption Recording and Summary: The system records the details of each resource consumption and displays the consumption history in a visual manner in the user's personal center.
[0006] Further technical solutions: Virtual resource units will be categorized into "bamboo" or "energy" for individual users and "honeydrop" or "work order" for enterprise users. Billing levels will be dynamically adjusted, triggered by factors including a decrease in call success rate, the introduction of more economical alternative models, and temporary adjustments during operational activities.
[0007] Beneficial effects: Value matching – Billing aligns with the actual value provided by AI, allowing users to pay for value rather than characters; User-friendly experience – The concept of virtual resources replaces cold monetary figures, lowering the psychological barrier for users to pay; Tiered guidance – Free quotas guide users to complete their first experience and become paying users; Brand alignment – The names of virtual resources are consistent with the platform's brand style.
Claims
1. Claim 1: A dynamic billing method based on the differentiation of AI agent roles, characterized in that, Includes the following steps: Each AI agent role and each tool type is configured with a separate billing tier, determined based on a combination of computing power cost, value proposition, and usage frequency. The monetary value of each billing tier is converted into a uniform virtual resource unit, which uses a user-friendly name consistent with the AI platform's brand style. When a user initiates an operation that requires resources, it is determined whether the user is within the free quota. If the free quota is exceeded, the corresponding amount of virtual resources is deducted according to the billing strategy, and the user is informed in a contextually embedded manner in the response. A tiered free quota is configured for each user, allocated according to three dimensions: agent role, time period, and operation type. Detailed records of each resource consumption are recorded, and the consumption history is displayed visually in the user's personal center.
2. Claim 2: The method according to claim 1, characterized in that, The naming of the virtual resource units is differentiated according to the target user group: "Bamboo" or "Energy" is used for individual users, and "Honey Drop" or "Work Order" is used for enterprise users.
3. Claim 3: The method according to claim 1, characterized in that, The billing tiers can be dynamically adjusted, with triggering conditions including a drop in call success rate to a preset threshold, the introduction of a more economical alternative model, and temporary adjustments during operational activities.
4. Claim 4: The method according to claim 1, characterized in that, The method of informing users in a natural contextual embedding manner specifically involves embedding lightweight expressions such as "I used XX to check" or "The deep analysis just now called an advanced service" in the response to the tool call, without interrupting the conversation flow.
5. Claim 5: The method according to claim 1, characterized in that, The tiered free quota includes: free first conversation for each intelligent agent role, a fixed number of free tool calls per day, and more free lightweight operations than heavyweight operations.
6. Claim 6: A dynamic billing system based on AI agent role differentiation, characterized in that, include: The billing configuration module is used to configure the billing level for each intelligent agent role and tool type; the virtual resource conversion module is used to convert monetary value into a uniform virtual resource unit. The billing execution module is used to determine the billing conditions and execute the deduction when an operation is triggered; The free quota management module is used to allocate and manage users' free quotas according to multiple dimensions. The record display module is used to record resource consumption details and display them in a visual way.