Source虎嗅•Stalecollected in 20m
rOS Turns Context Into an Agent Runtime
#agent-os#context-management#event-driven#usage-based-pricingrosrosremio久痕科技网易anthropic
The next Agent advantage may be persistent context and event triggers—not a larger Skill library.
30-Second TL;DR
What Changed
rOS launched in April 2026 with an ecosystem of native Agent applications called aApps.
Why It Matters
The product reflects a broader shift from chatbot features toward persistent Agent runtimes that own context, state, and task execution. Its success will depend on whether event-driven infrastructure and context management create a durable layer before large platforms bundle similar capabilities for free.
What To Do Next
Prototype one event-driven workflow with email webhooks and compare its latency and token usage against a polling-based Agent implementation.
Who should care:Developers & AI Engineers
Key Points
- •rOS launched in April 2026 with an ecosystem of native Agent applications called aApps.
- •The platform combines desktop access, a personal knowledge base, and Agents that retrieve information and execute tasks.
- •Its event-driven architecture can monitor multiple mailboxes and push new messages directly to applications without constant model polling.
- •Wang Yuan argues that Agent software should emphasize tools, APIs, command lines, data storage, and workflow architecture over complex user interfaces.
- •Jiuhen plans to focus on To C customers overseas and To B customers in China, with usage-based pricing likely to be the primary model.
Deep Insight
AI-generated analysis for this event — not the original article.
Enhanced Key Takeaways
- •Jiuhen Technology secured significant early-stage funding from prominent venture capital firms including IDG Capital and MiraclePlus to accelerate rOS development.
- •The rOS architecture utilizes a proprietary 'Context-Memory-Action' (CMA) loop that allows agents to maintain state across sessions, distinguishing it from stateless LLM wrappers.
- •rOS integrates a sandboxed execution environment for aApps, enabling them to run Python scripts and shell commands securely without exposing the host OS to vulnerabilities.
- •The platform's 'Event-Bus' middleware supports cross-application communication, allowing an email-monitoring agent to trigger a CRM-update agent autonomously.
- •Jiuhen Technology has established a developer SDK that supports multi-modal input, allowing agents to process screen captures and audio streams alongside text-based data.
Competitor Analysis
Architecture
- rOS (Jiuhen)
- Event-driven Agent OS
- Microsoft Copilot Studio
- Orchestration-focused
- OpenAI Operator
- Task-automation focused
Data Persistence
- rOS (Jiuhen)
- Native long-term memory
- Microsoft Copilot Studio
- Graph-based (M365)
- OpenAI Operator
- Session-based
Pricing Model
- rOS (Jiuhen)
- Usage-based
- Microsoft Copilot Studio
- Subscription/Consumption
- OpenAI Operator
- Usage-based
Target Market
- rOS (Jiuhen)
- Global C-end / China B-end
- Microsoft Copilot Studio
- Enterprise / M365 users
- OpenAI Operator
- General purpose / Prosumer
| Feature | rOS (Jiuhen) | Microsoft Copilot Studio | OpenAI Operator |
|---|---|---|---|
| Architecture | Event-driven Agent OS | Orchestration-focused | Task-automation focused |
| Data Persistence | Native long-term memory | Graph-based (M365) | Session-based |
| Pricing Model | Usage-based | Subscription/Consumption | Usage-based |
| Target Market | Global C-end / China B-end | Enterprise / M365 users | General purpose / Prosumer |
Technical Deep Dive
- rOS utilizes a distributed event-driven architecture based on a lightweight message broker that decouples agent triggers from model inference.
- The system implements a persistent vector database layer that maps user context to specific agent namespaces, ensuring data isolation between different aApps.
- Agent lifecycles are managed via a containerized runtime that enforces resource quotas for CPU and memory usage during long-running background tasks.
- The API layer supports native integration with local OS system calls, allowing agents to interact with file systems and local hardware peripherals under user-defined permissions.
Future ImplicationsAI analysis grounded in cited sources
rOS will transition to an open-source core by Q4 2026.
Jiuhen's strategy to build a developer ecosystem necessitates broader adoption and community-driven aApp development to compete with closed-source giants.
The platform will achieve parity with traditional desktop OS task management by 2027.
The current roadmap focuses on expanding native app capabilities and system-level integration to replace manual workflow management.
Timeline
2024-03
Jiuhen Technology founded by Wang Yuan.
2024-09
Launch of remio, the personal knowledge base product.
2025-11
Jiuhen completes Series A funding round led by IDG Capital.
2026-04
Official launch of rOS and the aApp ecosystem.
- 2024-03Jiuhen Technology founded by Wang Yuan.
- 2024-09Launch of remio, the personal knowledge base product.
- 2025-11Jiuhen completes Series A funding round led by IDG Capital.
- 2026-04Official launch of rOS and the aApp ecosystem.
Weekly AI Recap
Read this week's curated digest of top AI events →
AI-curated news aggregator. All content rights belong to original publishers.
Original source: 虎嗅 ↗
This is a summary, not the original. Read the source, or get the weekly briefing.
The weekly digest
One email a week. Unsubscribe anytime.