Product FAQ

Purpose: source document for the AI customer service bot (Xiaozhu) knowledge base — uploaded via the admin console under “AI Bot → Knowledge Base” and cited by RAG Q&A. Only public product statements belong here (consistent with the official website); do not include code, architecture, or other internal technical implementation details. When pricing or other public statements change, update this file and re-upload it.

Product and Positioning

Q: What is IMCore? A: IMCore is a self-hosted (on-premise) instant messaging system covering three conversation types — group chat, live room, and 1v1 direct messages — with optional AI-powered customer service and enterprise management capabilities. It runs on your own servers, and chat data never leaves your infrastructure.

Q: How is it different from usage-billed cloud IM (SaaS)? A: Cloud IM bills by active user count — the more users, the higher the cost — and data passes through the vendor’s servers. IMCore is licensed by annual subscription or one-time perpetual purchase: cost is decoupled from user count, data sovereignty stays entirely in your hands, and intranet/offline environments are supported.

Editions and Pricing

Q: What editions are available? A: Four editions — Basic (free, single-node IM core); Professional at ¥9,800/year (cluster scaling, offline push, content moderation, message callbacks); Customer Service at ¥19,800/year (Professional + the full AI customer service suite); Enterprise at ¥29,800/year (Professional + admin console + super community groups + external sync, recommended). Q: Can I buy a perpetual license? A: Yes. Perpetual license prices: Professional ¥29,800, Customer Service ¥59,800, Enterprise ¥89,800 (one-time). After the purchase, an optional annual upgrade and maintenance service is available at 10% (of the perpetual price) per year. Q: What happens when my subscription expires? A: The service keeps running — there is no shutdown — but version upgrades and technical support stop after expiry. Q: Can I upgrade from a lower edition? A: Yes. Basic and Professional are the same program, so upgrading only requires a license change; moving to Enterprise/Customer Service swaps in the corresponding program with the price difference paid. Data and configuration are fully compatible.

Deployment and Delivery

Q: How is it delivered? A: A prebuilt Linux binary distribution package plus deployment documentation — unpack it and deploy, no development environment required. The package contains the service binaries, configuration templates, database initialization tools, and integration docs. Q: What are the server requirements? A: One Linux server (x86 / ARM / Loongson architectures all work, compatible with Kylin, UOS, and other China-domestic systems) plus a database and Redis; clustered deployment additionally requires a message queue. You can start on a single node and later scale smoothly into a multi-instance cluster. Q: Which database should I use? A: PostgreSQL is recommended — the Customer Service edition’s AI/RAG capabilities require PostgreSQL (with the pgvector extension), and read/write splitting and Xinchuang (China domestic IT stack) database adaptation are also smoothest on the PG family; environments with existing MySQL infrastructure are fully supported as well. Q: Is source code provided? A: Delivery is binary-only by default, without source code. Custom requirements can be met through paid custom development; source-code licensing can be discussed with sales. Q: Does it support Xinchuang environments? A: Yes. It runs on China-domestic chips such as Hygon/Zhaoxin, Kunpeng/Phytium, and Loongson, and on China-domestic operating systems such as Kylin and UOS.

Features

Q: Which message types are supported? A: Text, images, video, files, and voice; with replies, quotes, forwarding, editing, recall, emoji reactions, read receipts, delivery status, presence, and typing indicators. Q: Are audio/video calls supported? A: Yes — 1v1 and multi-party audio/video calls. Q: What can the AI customer service do? A: Automatic bot reception (word-by-word streaming replies), knowledge base Q&A, rule-based auto-escalation to human agents, agent takeover and hand-back, voice message recognition, usage statistics and auditing. It can connect to a variety of LLM services. Q: Is there an admin console? A: The Enterprise and Customer Service editions include a web admin console: user/group/room/message management, sensitive-word management, notification templates, role permissions, and operation audit logs; operations trend dashboards are exclusive to the Enterprise edition. Q: How does content moderation work? A: A built-in Chinese sensitive-word dictionary works out of the box, with three policies — block, replace, and alert. Entries can be added/removed and bulk-imported in the admin console, and moderation records are queryable.

Integration

Q: How do I integrate the client SDKs? A: SDKs are provided for JS/Web, React Native, WeChat Mini Program, uni-app, Electron, iOS (Swift), Android (Kotlin), and Flutter (Dart), with accompanying integration docs and API references; chat UIKits are provided for Web/React, Flutter, Android Compose, and iOS SwiftUI. Q: How do we connect it to our existing user system? A: Three options to choose from as needed — the simplest is to pass nickname and avatar along with the auth token at login; for real-time sync, call the OpenAPI from your business code to push user profiles and friendships; for existing data, configure scheduled bulk synchronization. Q: Can we migrate from RongCloud? A: Yes. The Enterprise edition provides RongCloud migration tooling that supports cloud migration of history messages, friendships, and user profiles.

Trial and Support

Q: Where can I try it? A: Try the online demo at demo.balala.im (no sign-up, just open and use); the Basic edition can be self-hosted for free. Q: What does technical support include? A: Ticket and WeChat-group support; the Enterprise/Customer Service editions include deployment assistance. Exact service levels are governed by the contract. Q: How do I contact sales? A: Email sales@balala.im, or scan the QR code on the “Contact Us” page of the official website balala.im to add us on WeChat.