The Problem
Teams evaluating OpenMax need to connect plan limits, channel support, memory, review controls, and deployment requirements to one real workflow.
The Solution
This guide explains account setup, current plan structure, AI employee capabilities, channels, memory, deployment options, and the checks a team should complete before production use.
The Result
You get a practical evaluation path: select a workflow, verify the current product terms, run a controlled pilot, and expand only after quality and recovery checks pass.
How is OpenMax different from other AI platforms like ChatGPT or OpenClaw?
Before vs After OpenMax
Before
- Information scattered across the website; slow evaluation process
- Technical details and differentiation advantages unclear
- Pricing and payment methods require searching multiple pages
- Open-source ecosystem and product matrix lack systematic introduction
After
- Seven-category FAQ covering everything from registration to open source
- Official technical architecture explanation of Zylos and HxA Connect
- Four-tier pricing and Stripe/crypto payment clearly presented in one place
- A plan, deployment, data-handling, and pilot checklist in one place
FAQ: Registration and Login
Open the current OpenMax product entry and follow the account options shown in the interface. Before subscribing, review the available workspace, channel, plan, and role settings, because sign-in and onboarding steps can change with product updates.
FAQ: Plans and Pricing
OpenMax currently presents two public Agent Cloud plans plus custom enterprise options. Air at $99/mo includes 1 AI employee instance, standard compute, essential skills, all channels, and email support. Pro at $449/mo adds advanced compute, full skills, all channels, and 24-hour priority support. Enterprise is custom for unlimited instances, deep integration, SLA, SSO/SAML, and larger organizations. On-Premise is custom for fully self-hosted or private-cloud deployment where data must stay inside the organization. Air and Pro include persistent memory, 24/7 availability, and scheduled tasks.
FAQ: Payment Methods
Current purchase options and billing terms appear in the product checkout. Enterprise and On-Premise pricing, payment schedules, support terms, renewal, and cancellation conditions should be confirmed in the applicable order form or contract.
FAQ: AI Employee Capabilities and Limits
OpenMax employees are powered by the Zylos open-source runtime and feature persistent memory, scheduled tasks, and multi-channel communication. Unlike solo AI models, OpenMax employees retain business context long-term through five-layer memory architecture, auto-execute tasks on defined schedules, and collaborate in real time via HxA Connect. Air and Pro include 1 AI employee instance each. Enterprise supports unlimited scale and deeper integrations, while On-Premise supports private-cloud or self-hosted deployment for strict data residency. OpenMax uses fully managed model selection.
Technical Architecture and Open-Source Ecosystem
OpenMax combines Agent Cloud for AI employee operations, Zylos for runtime and memory, and HxA components for human-agent and channel coordination. Before deployment, confirm which components, interfaces, export paths, and hosting options are included in the plan you are evaluating.
How to validate an OpenMax use case
Choose one repeated workflow with a clear owner and a manual baseline. Give the pilot representative inputs, include normal exceptions, and record output quality, reviewer edits, failed actions, escalation handling, and weekly maintenance. Expand only when the workflow remains understandable and recoverable.
Still Have Questions?
Share your scenario, preferred channels, and deployment requirements with the OpenMax team. We can help you evaluate plans and launch the first AI employee faster.
About OpenMax
Product scope
OpenMax is designed for AI digital employees that work across recurring tasks, approved context, channels, and human review. Match those capabilities to a specific workflow instead of evaluating the product from a generic demo.
Runtime and integrations
Confirm the runtime components, supported interfaces, connector ownership, export formats, and migration path included in the selected plan.
Data handling
Map where prompts, uploaded files, memory, logs, model requests, and backups travel. Confirm retention, deletion, residency, and model-training terms for the actual plan and deployment model.
Commercial terms
Review the current checkout or contract for billing, renewal, cancellation, support response, usage limits, and what happens to data and active instances when service ends.
Frequently Asked Questions
What languages does OpenMax support?
OpenMax currently supports Chinese and English. AI employees communicate and execute tasks in your chosen language.
Can I use OpenMax on mobile?
Yes. OpenMax deploys AI employees to Telegram, Lark, Slack, and other channels, all of which have mobile apps.
Does OpenMax support private deployment?
Yes. The Enterprise plan includes a private deployment option. Contact the sales team for a customized proposal.
How should a team evaluate OpenMax?
Use one representative workflow with approved inputs, named owners, review rules, and measurable acceptance criteria. Confirm current plan, channel, security, and deployment terms before expanding.
Technical Notes
Model Management
OpenMax employs a fully managed model selection system. Rather than exposing individual model names (which change frequently as new versions release), OpenMax automatically routes each task to the optimal model based on task complexity, context size, and required reasoning depth. This approach ensures users always benefit from the latest model capabilities without manual configuration. Enterprise customers can discuss specific model preferences during onboarding.
Zylos Five-Layer Memory Architecture
The five layers are designed as follows: Layer 1 (Episodic): Raw conversation history, time-stamped. Layer 2 (Semantic): Extracted facts and entity relationships. Layer 3 (Procedural): Learned workflows and task patterns. Layer 4 (Preference): User-specific formatting, tone, and output style preferences. Layer 5 (Meta): Self-reflective optimization of which memories to prioritize for which task types. This layered approach provides stronger long-term retention than flat memory architectures.
Security Architecture
Data in transit is encrypted via TLS 1.3. Data at rest uses AES-256 encryption. User-uploaded documents are isolated per account with strict access controls. API communications between Agent Cloud and deployment channels (Telegram, Lark, Slack) use OAuth 2.0 with token-based authentication. Enterprise private deployment runs entirely within the customer's VPC or on-premise environment.
Questions to settle before choosing OpenMax
Users and roles
Clarify who creates AI employees, who approves access, who reviews outputs, and who owns each deployed workflow.
Channels and systems
List the channels and business systems the workflow must use, including the exact read and write operations required.
Data and deployment
Confirm data sensitivity, identity requirements, retention, audit logs, private deployment needs, and the human approval boundary.
How to use this FAQ in a rollout
Before a pilot
Use the answers to define the first workflow, owner, plan, channels, data access, and acceptance criteria.
During a pilot
Record unresolved questions, permission gaps, unexpected behavior, support needs, and changes required before go-live.
Before go-live
Reconfirm current plan limits, support coverage, deployment path, security controls, escalation ownership, and rollback.
