1. Basic information
ai.com is a personal assistant-type platform with "AI agent" as the core keyword, positioned as an autonomous AI agent that allows users to create and use tasks on the web side. The platform is currently in the beta stage, providing an account system and a username mechanism, allowing users to choose independent identities for themselves and agents, and complete registration and use within the scope of availability and rules. Its service provider is disclosed in the Terms as a corporate entity as Mentat Forge, Inc. (registered in Nevada) and provides services to users on behalf of ai.com and its affiliates.
2. Product overview
Unlike traditional "answer questions" chat-based tools, ai.com emphasizes that agents can perform actions with user authorization, such as organizing and organizing work, sending messages, completing actions across different applications, building projects, etc. The platform's public narrative long-term vision is to build a decentralized network of autonomous agents that enable agents to self-improve and share improvements to enhance overall capabilities. Regarding its underlying implementation, network mechanism and specific technical details, the disclosure of public materials is limited, and some details are uncertain.
3. Core functions
1. Main functions
1. Agent creation and personalization
Users can create "personal agents" and submit requests through dialogue input, files/images, etc. Proxy outputs include answers, summaries, and actionable suggestions.
2. Cross-application execution and integration
After the user turns on the third-party integration, the agent can access data from the relevant service to complete the request, such as processing email, calendar, and other productivity-related tasks. The specific integrations supported, the capability limits of each integration, and regional differences vary by region/version, and the actual available features shall prevail.
3. Username and identity system
The platform provides a username reservation mechanism, emphasizing that usernames are "licensed to use" rather than sold, and may be reclaimed or reassigned due to non-payment, long-term inactivity, violations, or brand protection.
2. Technical characteristics and usage
ai.com emphasizes the privacy idea of "isolation and encryption" in its external description, and distinguishes between "account and management information" and "proxy interaction and integrated data" in its privacy policy. At the same time, the privacy policy mentions that it does not use user prompts or proxy output for internal model training, but also states that it will share input with third-party large model service providers to generate responses and is subject to the terms of the other party. Users therefore need to make trade-offs between authorization scope, shared boundaries, and third-party rules.
4. Pricing and versions
The terms mention that there are free tiers and paid subscription tiers (e.g., Free, Plus, Pro, etc.), and that recurring billing and usage tracking are adopted. However, the public page does not fully disclose the specific prices, quotas, and differences for each tier, and the relevant information is uncertain and may be adjusted with the beta stage.
5. Applicable scenarios and target groups
ai.com Mainly for individual users and lightweight team collaboration who want to upgrade "conversation" to "task execution", common scenarios include: schedule and to-do organization, cross-application information aggregation, workflow automation, personal information management, and prompt-driven repetitive operations. For users who require strict data boundaries or sensitive information handling, prioritize assessing the risks posed by third-party integrations and third-party model calls.
6. Precautions for use and compliance
The Platform Terms set age and regional restrictions, such as minimum age requirements, and list jurisdictions that are sanctioned or access restricted. Before connecting to third-party services, users should check the scope of permissions, adopt the principle of minimum authorization as much as possible, and pay attention to account security, payment method compliance, and manual review of output. The platform also reminds us that beta features may be changed or withdrawn at any time.
7. Frequently asked questions
What is the difference between Q:ai.com's AI agents and regular chatbots?
A:ai.com Emphasize the ability to perform actions and cross-app actions with user authorization, rather than just conversational answers. However, the scope of enforcement depends on integration, permissions, and version capabilities.
Q:ai.com Will the model be trained with my prompts?
A: Its privacy policy states that it does not use prompts or proxy outputs to train internal models, but it may share inputs with third-party large model service providers to generate responses, subject to relevant terms and settings.
Q:ai.com What data is read when connecting to integrations such as mailboxes, calendars, etc.?
A: The Privacy Policy distinguishes between "active entry" and "passive integration data", after you connect to a third-party service, the agent may access task-related information in the service to complete the request, it is recommended to check the scope of authorization first and be cautious about sharing sensitive content.
Can a Q:ai.com username be owned forever?
A: The terms tend to treat usernames as licensed use rather than asset sales, and may be reclaimed or reallocated due to inactivity, violations, or brand protection, among other reasons.
