Help when it is needed
On-demand chat keeps the experience ready without forcing every visitor into a conversation. Customers can open support when they need it, continue in a familiar widget, and reach a real operator or AI-assisted first response without leaving your site.
Keep the entry point present and easy to find while letting the visitor decide when to engage.
Answer questions during checkout, onboarding, account management, and general support without changing channels.
The chat window, transcripts, routing, and analytics all stay part of the same self-hosted deployment.
Give customers a familiar chat flow that starts quickly and stays readable on both larger screens and compact layouts.
Collect the essentials up front and move the conversation into a structured support flow that can be handled by AI, a person, or both.
On-demand support means the entry point is always there and the decision is always the visitor's. That only works if every branch of the interaction is handled - including the ones where nobody is available.
The launcher sits in the corner and does nothing until it is clicked. No timed pop-up, no interstitial. If you do want to open conversations first, that is a separate, deliberately configured feature.
Any individual page can leave the widget out entirely, so a checkout flow can stay completely undisturbed.
The widget reflects real agent presence, and optional operating hours let you say you are closed rather than showing an online badge nobody will answer.
When every agent is at capacity the visitor is told their place in the queue instead of watching a spinner.
With nobody online the widget offers a leave-a-message form. Submissions become missed chats, complete with the visitor's details and the page they were reading, and can be pushed to you as a daily email digest.
An AI assistant can answer from your knowledge base and hand over to a person when the question needs one, carrying the conversation across. It is disabled until you enable it and choose a model.
The widget reports honestly that nobody is available and offers a leave-a-message form instead of an empty chat window. The submission is stored as a missed chat with the visitor's details and the page they were on, and you can have a daily digest emailed to your support address.
Yes. A visual schedule editor sets when the widget shows as online. Without a schedule the widget simply follows agent availability, showing online whenever someone is actually there.
Their position in the queue - 'you're next in line' rather than an unexplained wait - so they can decide whether to hold on or leave a message.
Yes, and it is off by default. When enabled it answers from your knowledge base and hands the conversation to a human when it should, carrying the context across so the visitor does not repeat themselves.
Yes. The conversation is remembered in the browser, so a refresh, a new tab, or a laptop waking from sleep resumes the same session rather than starting a new one.
ZChat gives you the installable server, web dashboard, website widget, and desktop agent tools in one self-hosted product you buy once and keep. Run it on infrastructure you trust and connect AI only if and how you want it.
Deployment
Install on Windows or Linux, behind IIS or Nginx, in a VM, or in Docker if that fits your stack.
Commercial model
One-time purchase, perpetual license, and no monthly per-agent bill attached to growth.
AI Flexibility
Use Ollama locally or connect OpenAI and Anthropic with your own provider accounts.