NextOlive Chat turns your WordPress site into a customer-messaging hub. Every feature is free and fully functional — there are no locked features, usage limits, or license checks.

Live chat widget with customizable colors, position, and welcome/offline messages.
Agent inbox inside WP admin — reply to visitors, see status, take over from AI.
Optional AI auto-reply powered by OpenAI (you supply your own API key and model). Automatic hand-off to a human on request.
Visitor monitoring — see who is online and what page they are on.
Support tickets / offline messages — visitors can leave a message when no one is around; agents reply from WP admin and the customer is emailed automatically. Status and priority management included.
Knowledge base — write help articles in WordPress; visitors search them from the chat widget, and the AI can use them to answer accurately (grounding).
Proactive triggers — greet visitors based on the page they’re on, time on site, page views, or whether they’re new/returning; optionally auto-open the chat.
Departments — route conversations to teams; optional visitor department picker; filter the inbox by department.
Canned responses / shortcuts — save reusable replies and insert them in the inbox by typing “/shortcut”.
Reports & analytics — chats/messages/visitors/tickets volume, average rating, average first-response time, chats-per-day chart, message breakdown, rating distribution, top sources and agent activity.
External embed — paste one script tag on any website (no WordPress required); it connects back to your WP install’s REST API.
Email notifications for new chats and new tickets.
Self-contained real-time via REST polling — no third-party account required to run the chat.

External services

This plugin can connect to one third-party service, and only when you explicitly enable it. No data is sent unless you take the action described.

OpenAI API (https://api.openai.com)

Used only if you enable “AI auto-reply” and enter your own OpenAI API key under NextOlive Chat Settings. When AI auto-reply is enabled and a visitor sends a chat message, the conversation text and your configured system prompt (and, if you enable knowledge-base grounding, excerpts of your published KB articles) are sent to OpenAI so it can generate a reply. Nothing is sent to OpenAI while AI auto-reply is disabled or no API key is set.

No other external services are contacted.