內容簡介
AI Chatbot Free Models 是一個完整的 AI 聊天機器人平台，專為 WordPress 設計。它提供專業的即時聊天、客戶支援和虛擬助手體驗，並讓使用者自由選擇 AI 模型，無需訂閱或外部帳號。
【主要功能】
• 提供者無關的 AI 連接，支持多種端點
• 動態模型選擇，支持模型發現
• 實時回應與輸入指示器
• 可選的預聊天表單收集資訊
• 實時知識庫，支持全文索引搜索
• 浮動小工具或嵌入式聊天選項
外掛標籤
開發者團隊
② 後台搜尋「AI Chatbot Free Models – Customer Support, Live Chat, Virtual Assistant」→ 直接安裝（推薦）
原文外掛簡介
AI Chatbot Free Models is a fully self-contained AI chatbot platform for WordPress. It provides everything you need to run professional live chat, customer support and virtual-assistant experiences on your website while leaving the choice of the AI model entirely in your hands.
There is no paid SaaS behind this plugin. No mandatory subscription, no premium gateway, no external dashboard, no account required with the plugin developer. The only thing you need is access to an AI model: a hosted provider (many of which offer generous free tiers) or a self-hosted, local inference server. Every other component – interface, conversations, knowledge base, notifications, styling – runs on your own WordPress installation and your own database.
Bring your own model, keep control of your data
Connect OpenRouter, OpenAI, Anthropic, Google or any OpenAI-compatible endpoint.
Run completely free with free-tier models or a local server (llama.cpp, Ollama, LM Studio, vLLM, gateways and proxies).
Conversations are stored only if you enable it, in dedicated tables inside your own database.
The knowledge base is built from your own published posts and pages – no external crawling, no third-party index.
Core features
Provider-agnostic AI connection with configurable endpoint, API key and model
Dynamic model selection where the provider supports model discovery
Streaming responses with live typing indicator
Stop generation at any time while a response is being streamed
Retry failed responses and start a new conversation without reloading the page
Configurable maximum response tokens and streaming timeout with inactivity watchdog
Conversation context maintained with a configurable token budget
Optional pre-chat form to collect name and email for support and lead generation
Live knowledge base with FULLTEXT-indexed search over titles and content, response caching and clickable source citations in answers
Per-IP rate limiting and server-side sanitization of conversation history
Dedicated database storage with optional message and IP retention (disabled by default)
Unread conversation notifications, search and conversation management in the WordPress admin
Floating widget (bottom-left or bottom-right) or embedded chat via the [newcodebyte_chatbot] shortcode
Extensive appearance customization: colors, images, sounds, widget position
Translatable frontend interface, responsive layout, multilingual AI support
Modern object-oriented codebase following current WordPress standards
Requirements
WordPress 5.0 or later (tested up to 7.1)
PHP 7.4 or later (compatible with current PHP 8.x releases)
MySQL 5.6 or later / MariaDB 10.0 or later (required by the FULLTEXT knowledge base index)
An API key for the provider you choose – or none at all, if you use a local OpenAI-compatible server
How it works
A visitor sends a message from the floating widget or the embedded chat.
WordPress receives the request and the plugin prepares the conversation context.
The configured provider or endpoint is selected and the request is sent.
The response is streamed back to the visitor when streaming is supported.
The conversation can optionally be stored in your own database for review.
A note on costs
The plugin is free and adds no usage fees of its own. AI costs depend exclusively on the provider and model you configure: free models and free quotas are fully supported, and a self-hosted server removes usage costs entirely.
Looking for a platform-independent version? A standalone build for non-WordPress websites (HTML, PHP and more) is available on GitHub: https://github.com/NewCodeByte/chatbot-ai-free-models
Configuration
Open the Chatbot settings page. Depending on the selected provider you can configure:
Provider (OpenRouter, OpenAI, Anthropic, Google or OpenAI-compatible endpoint)
API endpoint and API key
Model, selected from discovery or entered manually
Maximum response tokens and streaming timeout
Knowledge base, pre-chat form, storage options, appearance and sounds
Free models on OpenRouter are identified by the :free suffix. When using a local or self-hosted OpenAI-compatible server, simply point the endpoint to your server (for example http://localhost:8080/v1): no external account is required and your traffic never leaves your infrastructure.
External Services
The plugin can transmit the content of a visitor’s message to the AI provider you configure (OpenRouter, OpenAI, Anthropic, Google or a custom endpoint) in order to generate a response. The terms of service and privacy policy of the selected provider apply to that processing. The plugin developer does not control and is not responsible for external providers’ pricing, availability, retention policies or data processing.
Privacy
Message and IP storage are optional and disabled by default.
When enabled, data is stored in dedicated tables inside your own database and can be reviewed, exported or deleted from the admin area.
The plugin itself sets no cookies.
If you enable storage or the pre-chat form, update your privacy policy to describe the processing. You are responsible for compliance with applicable data-protection laws, including GDPR.
Disclaimer
AI-generated content can be inaccurate. The plugin developer is not responsible for AI responses, decisions made based on them, provider outages, pricing or API changes, or data processing performed by external providers. Always verify important information independently.
Supporting the project
AI Chatbot Free Models is developed and maintained as an independent free project. If you find it useful, consider keeping the optional “Powered by NewCodeByte” attribution enabled or supporting development with a donation. Contributions fund development, maintenance, testing and infrastructure.
