diff --git a/i18n/en/docusaurus-plugin-content-docs/version-5.0/01-introduction/01whychoose.md b/i18n/en/docusaurus-plugin-content-docs/version-5.0/01-introduction/01whychoose.md
index 89767acb3e..0a6a40466f 100644
--- a/i18n/en/docusaurus-plugin-content-docs/version-5.0/01-introduction/01whychoose.md
+++ b/i18n/en/docusaurus-plugin-content-docs/version-5.0/01-introduction/01whychoose.md
@@ -2,25 +2,98 @@
slug: /
---
-# Why choose RocketMQ
+# Why Choose RocketMQ
-## Why RocketMQ
-During the early stages of RocketMQ's development at Alibaba, we utilized it for a multitude of purposes, including asynchronous communications, search, social networking activity flows, data pipelines, and trade processes. As our trade business grew, we noticed that the messaging cluster was under increasing pressure.
+## AI-Native Asynchronous Communication Engine
-After observing and analyzing the performance of the ActiveMQ IO module, we identified a bottleneck as the number of queues and virtual topics increased. We attempted to address this issue through various methods, such as throttling, circuit breakers, and service downgrades, but none proved satisfactory. We also considered using Kafka, a popular messaging solution, but it did not meet our requirements for low latency and high reliability, as explained below. As a result, we made the decision to develop a new messaging engine capable of handling a wider range of use cases, from traditional pub/sub to high-volume, real-time, zero-error transaction systems.
+Apache RocketMQ is an AI-native asynchronous communication engine purpose-built for trillion-message scale. Born from Alibaba's high-concurrency e-commerce workloads and battle-tested across thousands of enterprises, RocketMQ has evolved from a high-performance message queue into a unified messaging platform that bridges three paradigms: traditional business messaging, event streaming, and the emerging AI-native communication.
-Since its inception, Apache RocketMQ has been widely adopted by enterprise developers and cloud vendors due to its simple architecture, rich business functionality, and extreme scalability. After more than a decade of extensive scenario polishing, RocketMQ has become the industry standard for financial-grade reliable business messages and is widely used in Internet, big data, mobile Internet, IoT, and other fields.
+## Core Strengths
-:::tip
+### AI-Native Capabilities (New in 5.5)
-The following table shows the comparison between RocketMQ, ActiveMQ and Kafka
+RocketMQ 5.5.0 (released April 2026) introduces a strategic upgrade for AI workloads:
-:::
+- **Million-Scale LiteTopics**: Purpose-built for AI Agent session management, LiteTopics enable millions of lightweight channels with minimal resource overhead. Each AI Agent conversation session maps to an independent topic, delivering fine-grained state isolation and lifecycle management powered by RocksDB indexing.
+- **Lite Mode Subscription**: A lightweight subscription model designed for AI scenarios, offering efficient message passing with significantly lower resource consumption — ideal for event-driven agent orchestration where thousands of agents need real-time coordination.
+- **Intelligent Compute Scheduling**: Solves critical AI workflow challenges including long-running session drops, resource waste from idle connections, and cascading blocks in multi-agent pipelines through event-driven pull mechanisms.
-## RocketMQ vs. ActiveMQ vs. Kafka
+### Cloud-Native Architecture
-| Messaging Product|Client SDK| Protocol and Specification | Ordered Message | Scheduled Message | Batched Message |BroadCast Message| Message Filter|Server Triggered Redelivery|Message Storage|Message Retroactive|Message Priority|High Availability and Failover|Message Track|Configuration|Management and Operation Tools|
-| -------|--------|--------|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
-| ActiveMQ|Java, .NET, C++ etc. |Push model, support OpenWire, STOMP, AMQP, MQTT, JMS|Exclusive Consumer or Exclusive Queues can ensure ordering|Supported|Not Supported|Supported|Supported|Not Supported|Supports very fast persistence using JDBC along with a high performance journal，such as levelDB, kahaDB|Supported|Supported|Supported, depending on storage,if using levelDB it requires a ZooKeeper server|Not Supported|The default configuration is low level, user need to optimize the configuration parameters|Supported|
-| Kafka | Java, Scala etc.|Pull model, support TCP|Ensure ordering of messages within a partition|Not Supported|Supported, with async producer|Not Supported|Supported, you can use Kafka Streams to filter messages|Not Supported|High performance file storage|Supported offset indicate|Not Supported|Supported, requires a ZooKeeper server|Not Supported|Kafka uses key-value pairs format for configuration. These values can be supplied either from a file or programmatically.|Supported, use terminal command to expose core metrics|
-| RocketMQ |Java, C++, Go |Pull model, support TCP, JMS, OpenMessaging|Ensure strict ordering of messages,and can scale out gracefully|Supported|Supported, with sync mode to avoid message loss|Supported|Supported, property filter expressions based on SQL92|Supported|High performance and low latency file storage|Supported timestamp and offset two indicates|Not Supported|Supported, Master-Slave model, without another kit|Supported|Work out of box,user only need to pay attention to a few configurations|Supported, rich web and terminal command to expose core metrics|
+RocketMQ 5.x embraces a fully cloud-native design:
+
+- **Elastic Scaling**: Stateless proxy layer and separated compute-storage architecture enable independent scaling of throughput and storage capacity.
+- **Kubernetes Native**: First-class support for container orchestration with Helm charts, Operators, and seamless integration with cloud-native observability stacks.
+- **Lightweight gRPC SDKs**: Multi-language clients (Java, Go, C++, Rust, Python, Node.js) built on gRPC provide consistent APIs and lower integration complexity.
+- **Multi-Protocol Support**: Native support for gRPC, MQTT, AMQP, and HTTP/REST enables unified messaging across IoT devices, microservices, and serverless functions.
+
+### Financial-Grade Reliability
+
+RocketMQ is the industry standard for financial-grade reliable business messaging. It provides:
+
+- **Transactional Messages**: Native two-phase commit ensures exactly-once delivery semantics across distributed systems, allowing message sending and local database transactions to succeed or fail atomically.
+- **High Availability**: DLedger-based Raft consensus ensures zero message loss with automatic leader election and multi-replica synchronization.
+- **Strict Message Ordering**: FIFO ordering guarantees at partition level support business scenarios that demand sequential processing — order creation, payment, and fulfillment can be strictly serialized.
+
+## Messaging Patterns
+
+RocketMQ supports a rich set of messaging patterns within a single platform:
+
+- **Publish/Subscribe**: Classic topic-based fan-out for decoupled microservice communication.
+- **Request/Reply**: Synchronous-style RPC over asynchronous messaging for service invocation patterns.
+- **Event Streaming**: Continuous event processing with consumer offsets, replay capabilities, and integration with stream processing frameworks.
+- **Delay & Scheduled Messages**: Arbitrary-precision scheduled delivery for business workflows like order timeout cancellation and retry scheduling.
+- **Batch Consumption**: High-throughput batch pull for data pipeline and analytics scenarios.
+
+## Feature Comparison
+
+| Feature | RocketMQ | ActiveMQ | Kafka |
+|---------|----------|----------|-------|
+| AI Agent Communication | LiteTopic | Not supported | Not supported |
+| Message Ordering | Strict FIFO per partition | Exclusive consumer | Partition-level only |
+| Message Filtering | SQL92 + Tag-based | Limited | Topic-level only |
+| Transactional Message | Native 2PC | XA (heavy) | Not supported |
+| Delayed Message | Arbitrary precision | Limited levels | Not supported |
+| Dead Letter Queue | Built-in with retry | Built-in | Manual implementation |
+| Message Tracing | Built-in full-link tracing | Plugin-based | Not supported |
+| Horizontal Scaling | Linear scale-out | Limited | Partition-bound |
+| Operational Tooling | Rich dashboard & CLI | Basic web console | Third-party tools |
+| Protocol Support | gRPC/MQTT/AMQP/HTTP | AMQP/STOMP/MQTT | Custom TCP |
+
+## Use Cases
+
+### AI Agent Communication
+
+The newest frontier — RocketMQ provides the asynchronous communication infrastructure for multi-agent AI systems. Aligned with the AI-native vision, RocketMQ delivers four scenario-based solutions:
+
+- **MCP & Long-Session State Continuity**: Each session maps to a dedicated LiteTopic, keeping application servers stateless. On client reconnection, sessions resume from breakpoint while backend LLM tasks continue running uninterrupted — eliminating session loss and wasted compute in long-running conversations.
+- **Fine-Grained Isolation & Rate Limiting**: Per-LiteTopic Suspend/Resume at the consumer level enables millisecond-level rate limiting. Anomalies in a single session or agent are isolated without cascading impact, delivering smooth traffic curves and elastic scaling for multi-tenant, multi-agent deployments.
+- **Multi-Agent Async Communication**: Native support for MCP and A2A protocols, with seamless integration into mainstream agent frameworks including LangChain, CrewAI, AutoGen, AgentScope, Dify, and Coze — and full compatibility with framework-agnostic systems. Event-driven async messaging replaces static orchestration, enabling more flexible and efficient agent collaboration.
+- **Batch Consumption — Unleash GPU Parallelism**: Dual-trigger batching by message count and time window. Both PushConsumer and SimpleConsumer modes adapt to LLM Batch API patterns with adaptive pacing under varying throughput, maximizing GPU inference throughput and parallel compute utilization.
+
+### Business Messaging & Microservices
+
+RocketMQ serves as the asynchronous communication backbone for microservice architectures, handling order processing, payment notifications, inventory synchronization, and cross-service event propagation with financial-grade transactional guarantees.
+
+### Event Streaming & Data Integration
+
+With built-in RocketMQ Connect, the platform integrates with 30+ data sources and sinks (MySQL, PostgreSQL, Elasticsearch, Hudi, etc.), enabling real-time CDC pipelines, data lake ingestion, and cross-system event sourcing.
+
+### IoT & Edge Computing
+
+MQTT protocol support with million-level device connections makes RocketMQ suitable for IoT telemetry collection, device command dispatch, and edge-cloud message bridging scenarios.
+
+## Ecosystem & Community
+
+Apache RocketMQ benefits from a vibrant open-source community:
+
+- **75+ Contributors** with continuous active development on GitHub
+- **RocketMQ Streams**: Lightweight stream processing library for real-time computation
+- **RocketMQ Connect**: Standardized connector framework for data integration
+- **RocketMQ MQTT**: Protocol bridge for IoT workloads
+- **RocketMQ Dashboard**: Full-featured web console for operations and monitoring
+- **Multi-Cloud Deployment**: Available as managed service on major cloud platforms (Alibaba Cloud, AWS, Huawei Cloud)
+
+---
+
+*Apache RocketMQ — from battle-tested business messaging to AI-native asynchronous communication, powering the next generation of distributed systems.*
diff --git a/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/02ratelimit.md b/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/02ratelimit.md
index f1f77fd4c0..c300a216e9 100644
--- a/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/02ratelimit.md
+++ b/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/02ratelimit.md
@@ -27,7 +27,7 @@ Kafka faces the same heavyweight resource issues as the RocketMQ standard Topic
## Solution
### Core Architecture: Fine-Grained Isolation and Rate Limiting Based on LiteTopic
-
+
This document adopts Apache RocketMQ's **LiteTopic (Lightweight Topic)** feature to build a novel fine-grained isolation and dynamic rate limiting architecture:
diff --git a/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/04multiagent.md b/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/04multiagent.md
index 7776b4055f..609f40596e 100644
--- a/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/04multiagent.md
+++ b/i18n/en/docusaurus-plugin-content-docs/version-5.0/06-bestPractice/04multiagent.md
@@ -10,7 +10,7 @@ As AI applications become increasingly complex, single-agent architectures are o
### System architecture
-
+
* **Web client**: The user interface for sending requests and receiving final results.
diff --git a/i18n/en/docusaurus-plugin-content-docs/version-5.0/picture/v5/litetopic_multiagent_practice.png b/i18n/en/docusaurus-plugin-content-docs/version-5.0/picture/v5/litetopic_multiagent_practice.png
index 0b02f2e010..2637d95cc1 100644
Binary files a/i18n/en/docusaurus-plugin-content-docs/version-5.0/picture/v5/litetopic_multiagent_practice.png and b/i18n/en/docusaurus-plugin-content-docs/version-5.0/picture/v5/litetopic_multiagent_practice.png differ
diff --git a/i18n/en/docusaurus-plugin-content-docs/version-5.0/picture/v5/rocketmq-litetopic-rate-limit-architecture-en.png b/i18n/en/docusaurus-plugin-content-docs/version-5.0/picture/v5/rocketmq-litetopic-rate-limit-architecture-en.png
new file mode 100644
index 0000000000..1a4fac6395
Binary files /dev/null and b/i18n/en/docusaurus-plugin-content-docs/version-5.0/picture/v5/rocketmq-litetopic-rate-limit-architecture-en.png differ
diff --git a/src/pages/components/AISolutions.js b/src/pages/components/AISolutions.js
index 68b9633442..3a99507a1b 100644
--- a/src/pages/components/AISolutions.js
+++ b/src/pages/components/AISolutions.js
@@ -81,7 +81,7 @@ function RateLimitPanel() {
@@ -136,7 +136,7 @@ function MultiAgentPanel() {
diff --git a/static/picture/v5/litetopic_multiagent_practice.png b/static/picture/v5/litetopic_multiagent_practice.png
new file mode 100644
index 0000000000..2637d95cc1
Binary files /dev/null and b/static/picture/v5/litetopic_multiagent_practice.png differ
diff --git a/static/picture/v5/litetopic_multiagent_practice.svg b/static/picture/v5/litetopic_multiagent_practice.svg
index f482a31af7..a2ed562552 100644
--- a/static/picture/v5/litetopic_multiagent_practice.svg
+++ b/static/picture/v5/litetopic_multiagent_practice.svg
@@ -1 +1,3 @@
-
\ No newline at end of file
+
+
+
\ No newline at end of file
diff --git a/static/picture/v5/litetopic_session_practice.svg b/static/picture/v5/litetopic_session_practice.svg
index c50997897b..82a102af86 100644
--- a/static/picture/v5/litetopic_session_practice.svg
+++ b/static/picture/v5/litetopic_session_practice.svg
@@ -1 +1,3 @@
-
\ No newline at end of file
+
+
+
\ No newline at end of file
diff --git a/static/picture/v5/rocketmq-litetopic-rate-limit-architecture-en.png b/static/picture/v5/rocketmq-litetopic-rate-limit-architecture-en.png
new file mode 100644
index 0000000000..1a4fac6395
Binary files /dev/null and b/static/picture/v5/rocketmq-litetopic-rate-limit-architecture-en.png differ
diff --git a/versioned_docs/version-5.0/01-introduction/01whychoose.md b/versioned_docs/version-5.0/01-introduction/01whychoose.md
index 365d868bfb..8dbfae3d6c 100644
--- a/versioned_docs/version-5.0/01-introduction/01whychoose.md
+++ b/versioned_docs/version-5.0/01-introduction/01whychoose.md
@@ -2,25 +2,99 @@
slug: /
---
-# 为什么选择RocketMQ
+# 为什么选择 RocketMQ
-## 为什么 RocketMQ
-在阿里孕育 RocketMQ 的雏形时期，我们将其用于异步通信、搜索、社交网络活动流、数据管道，贸易流程中。随着我们的贸易业务吞吐量的上升，源自我们的消息传递集群的压力也变得紧迫。
+## AI 原生异步通信引擎
-根据我们的研究，随着队列和虚拟主题使用的增加，ActiveMQ IO模块达到了一个瓶颈。我们尽力通过节流、断路器或降级来解决这个问题，但效果并不理想。于是我们尝试了流行的消息传递解决方案Kafka。不幸的是，Kafka不能满足我们的要求，其尤其表现在低延迟和高可靠性方面，详见下文。在这种情况下，我们决定发明一个新的消息传递引擎来处理更广泛的消息用例，覆盖从传统的pub/sub场景到高容量的实时零误差的交易系统。
+Apache RocketMQ 是一款面向万亿级消息规模的 AI 原生异步通信引擎。诞生于阿里巴巴高并发电商场景，经过数千家企业的生产验证，RocketMQ 已从高性能消息队列演进为统一消息平台，横跨传统业务消息、事件流处理和新兴的 AI 原生通信三大范式。
-Apache RocketMQ 自诞生以来，因其架构简单、业务功能丰富、具备极强可扩展性等特点被众多企业开发者以及云厂商广泛采用。历经十余年的大规模场景打磨，RocketMQ 已经成为业内共识的金融级可靠业务消息首选方案，被广泛应用于互联网、大数据、移动互联网、物联网等领域的业务场景。
+## 核心优势
-:::tip
+### AI 原生能力（5.5 新特性）
-下表显示了RocketMQ、ActiveMQ和Kafka之间的比较
+RocketMQ 5.5.0（2026年4月发布）引入面向 AI 工作负载的战略升级：
-:::
+- **百万级 LiteTopic**：专为 AI Agent 会话管理设计，以极低资源开销支持百万级轻量通道。每个 AI Agent 对话会话映射为独立 Topic，基于 RocksDB 索引实现细粒度状态隔离与生命周期管理。
+- **Lite Mode 轻量订阅**：面向 AI 场景的轻量级订阅模型，以显著更低的资源消耗实现高效消息传递——适用于数千 Agent 需要实时协调的事件驱动编排场景。
+- **智能计算调度**：通过事件驱动拉取机制，解决 AI 工作流中的长会话断连、空闲连接资源浪费、多 Agent 管线级联阻塞等关键问题。
-## RocketMQ vs. ActiveMQ vs. Kafka
+### 云原生架构
-| Messaging Product|Client SDK| Protocol and Specification | Ordered Message | Scheduled Message | Batched Message |BroadCast Message| Message Filter|Server Triggered Redelivery|Message Storage|Message Retroactive|Message Priority|High Availability and Failover|Message Track|Configuration|Management and Operation Tools|
-| -------|--------|--------|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
-| ActiveMQ|Java, .NET, C++ etc. |Push model, support OpenWire, STOMP, AMQP, MQTT, JMS|Exclusive Consumer or Exclusive Queues can ensure ordering|Supported|Not Supported|Supported|Supported|Not Supported|Supports very fast persistence using JDBC along with a high performance journal，such as levelDB, kahaDB|Supported|Supported|Supported, depending on storage,if using levelDB it requires a ZooKeeper server|Not Supported|The default configuration is low level, user need to optimize the configuration parameters|Supported|
-| Kafka | Java, Scala etc.|Pull model, support TCP|Ensure ordering of messages within a partition|Not Supported|Supported, with async producer|Not Supported|Supported, you can use Kafka Streams to filter messages|Not Supported|High performance file storage|Supported offset indicate|Not Supported|Supported, requires a ZooKeeper server|Not Supported|Kafka uses key-value pairs format for configuration. These values can be supplied either from a file or programmatically.|Supported, use terminal command to expose core metrics|
-| RocketMQ |Java, C++, Go |Pull model, support TCP, JMS, OpenMessaging|Ensure strict ordering of messages,and can scale out gracefully|Supported|Supported, with sync mode to avoid message loss|Supported|Supported, property filter expressions based on SQL92|Supported|High performance and low latency file storage|Supported timestamp and offset two indicates|Not Supported|Supported, Master-Slave model, without another kit|Supported|Work out of box,user only need to pay attention to a few configurations|Supported, rich web and terminal command to expose core metrics|
+RocketMQ 5.x 采用全面云原生设计：
+
+- **弹性伸缩**：无状态 Proxy 层与存算分离架构，支持吞吐与存储容量独立扩缩。
+- **Kubernetes 原生**：提供 Helm Charts、Operator 以及与云原生可观测体系的无缝集成。
+- **轻量 gRPC SDK**：多语言客户端（Java、Go、C++、Rust、Python、Node.js）基于 gRPC 构建，提供一致的 API 和更低的集成复杂度。
+- **多协议支持**：原生支持 gRPC、MQTT、AMQP 和 HTTP/REST，统一覆盖 IoT 设备、微服务和 Serverless 函数的消息需求。
+
+### 金融级可靠性
+
+RocketMQ 是金融级可靠业务消息的行业标准，提供：
+
+- **事务消息**：原生两阶段提交保证分布式系统中的精确一次投递语义，消息发送与本地数据库事务原子性地同时成功或失败。
+- **高可用架构**：基于 DLedger 的 Raft 一致性协议确保消息零丢失，支持自动选主和多副本同步复制。
+- **严格顺序消息**：分区级 FIFO 顺序保证，支撑订单创建、支付扣款、库存扣减等要求严格串行化处理的业务场景。
+
+## 消息模式
+
+RocketMQ 在单一平台内支持丰富的消息模式：
+
+- **发布/订阅**：经典的基于 Topic 的扇出模型，实现微服务间解耦通信。
+- **请求/应答**：基于异步消息的同步风格 RPC，用于服务调用场景。
+- **事件流**：支持消费位点管理、消息回溯、与流处理框架集成的持续事件处理。
+- **延迟与定时消息**：任意精度的定时投递，支撑订单超时取消、重试调度等业务流程。
+- **批量消费**：高吞吐批量拉取，面向数据管道和分析场景。
+
+## 功能对比
+
+| 特性 | RocketMQ | ActiveMQ | Kafka |
+|------|----------|----------|-------|
+| AI Agent通信 | LiteTopic | 不支持 | 不支持 |
+| 消息顺序 | 分区级严格 FIFO | 独占消费者 | 仅分区级 |
+| 消息过滤 | SQL92 + Tag 过滤 | 有限支持 | 仅 Topic 级 |
+| 事务消息 | 原生两阶段提交 | XA（重量级） | 不支持 |
+| 延迟消息 | 任意精度 | 有限级别 | 不支持 |
+| 死信队列 | 内置重试机制 | 内置 | 需手动实现 |
+| 消息轨迹 | 内置全链路追踪 | 插件实现 | 不支持 |
+| 水平扩展 | 线性扩展 | 有限 | 受分区数约束 |
+| 运维工具 | 丰富的 Dashboard 与 CLI | 基础 Web 控制台 | 依赖第三方工具 |
+| 协议支持 | gRPC/MQTT/AMQP/HTTP | AMQP/STOMP/MQTT | 自定义 TCP |
+
+
+## 典型场景
+
+### AI Agent 通信
+
+RocketMQ 的最新前沿方向——为多 Agent AI 系统提供异步通信基础设施。围绕 AI 原生场景，RocketMQ 提供四大场景解决方案：
+
+- **MCP & 长会话状态续连**：每个会话映射到一个独立的 LiteTopic，应用服务器保持无状态。客户端断线重连后可断点续传，后端 LLM 推理任务持续运行不中断，彻底解决长对话场景下的会话丢失与算力浪费问题。
+- **细粒度隔离与限流**：在消费者侧支持以 LiteTopic 为粒度的 Suspend/Resume 操作，实现毫秒级精细化限流。单个会话或 Agent 的异常不会扩散，配合平滑的流量曲线与弹性扩缩，保障多租户、多 Agent 共存场景下的稳定性。
+- **多 Agent 异步通信**：原生支持 MCP 与 A2A 协议，无缝对接 LangChain、CrewAI、AutoGen、AgentScope、Dify、Coze 等主流 Agent 框架，同时也兼容自研框架。事件驱动的异步通信替代静态编排，让 Agent 协作更灵活、更高效。
+- **批量消费——释放 GPU 并行算力**：通过消息条数 + 时间窗口双触发机制，PushConsumer 与 SimpleConsumer 模式都可以适配 LLM Batch API 的调用范式，自适应不同吞吐压力下的批处理节奏，最大化 GPU 推理吞吐与利用率。
+
+### 业务消息与微服务
+
+RocketMQ 作为微服务架构的异步通信骨干，处理订单处理、支付通知、库存同步和跨服务事件传播，提供金融级事务保证。
+
+### 事件流与数据集成
+
+通过内置的 RocketMQ Connect，平台集成 30+ 数据源和目标端（MySQL、PostgreSQL、Elasticsearch、Hudi 等），支持实时 CDC 管道、数据湖入湖和跨系统事件溯源。
+
+### IoT 与边缘计算
+
+MQTT 协议支持百万级设备连接，适用于 IoT 遥测采集、设备指令下发和边云消息桥接场景。
+
+## 生态与社区
+
+Apache RocketMQ 拥有活跃的开源社区：
+
+- **75+ 贡献者**，GitHub 上持续活跃开发
+- **RocketMQ Streams**：轻量级流处理库，支持实时计算
+- **RocketMQ Connect**：标准化连接器框架，用于数据集成
+- **RocketMQ MQTT**：IoT 工作负载的协议桥接
+- **RocketMQ Dashboard**：功能完备的 Web 运维控制台
+- **多云部署**：在主流云平台提供托管服务（阿里云、AWS、华为云）
+
+---
+
+*Apache RocketMQ —— 从久经考验的业务消息到 AI 原生异步通信，驱动下一代分布式系统。*
diff --git a/versioned_docs/version-5.0/06-bestPractice/02ratelimit.md b/versioned_docs/version-5.0/06-bestPractice/02ratelimit.md
index 6147022f38..129194f70d 100644
--- a/versioned_docs/version-5.0/06-bestPractice/02ratelimit.md
+++ b/versioned_docs/version-5.0/06-bestPractice/02ratelimit.md
@@ -27,7 +27,7 @@ Kafka 与 RocketMQ 普通 Topic 方案面临相同的重资源问题。若尝试
## 解决方案
### 核心架构：基于LiteTopic的细粒度隔离和限流
-
+
本文采用 Apache RocketMQ 的 **LiteTopic（轻量级主题）** 特性，构建了一套全新的细粒度隔离与动态限流架构：
diff --git a/versioned_docs/version-5.0/06-bestPractice/04multiagent.md b/versioned_docs/version-5.0/06-bestPractice/04multiagent.md
index 31ace18f49..a5654d99b9 100644
--- a/versioned_docs/version-5.0/06-bestPractice/04multiagent.md
+++ b/versioned_docs/version-5.0/06-bestPractice/04multiagent.md
@@ -10,7 +10,7 @@
### 系统架构
-
+
* **Web端**：用户交互界面，向系统发起请求并接收最终结果。
diff --git a/versioned_docs/version-5.0/picture/v5/litetopic_multiagent_practice.png b/versioned_docs/version-5.0/picture/v5/litetopic_multiagent_practice.png
index 0b02f2e010..28854b6c16 100644
Binary files a/versioned_docs/version-5.0/picture/v5/litetopic_multiagent_practice.png and b/versioned_docs/version-5.0/picture/v5/litetopic_multiagent_practice.png differ
diff --git a/versioned_docs/version-5.0/picture/v5/rocketmq-litetopic-rate-limit-architecture.png b/versioned_docs/version-5.0/picture/v5/rocketmq-litetopic-rate-limit-architecture.png
new file mode 100644
index 0000000000..a7779ccdca
Binary files /dev/null and b/versioned_docs/version-5.0/picture/v5/rocketmq-litetopic-rate-limit-architecture.png differ