AskAny – AI Chatbot with RAG, Live Agent and Order Tracking vs AxiaChat AI – Free AI Chatbot (Answers Customers Automatically)

AskAny is a free AI chatbot for WordPress built on RAG (Retrieval-Augmented Generation). Before it answers anything, it retrieves the most relevant sections of your actual site content using semantic search, then generates a grounded, cited answer from that content instead of guessing from general AI knowledge. Several other WordPress AI chatbot plugins, including AI Engine and MxChat, also use retrieval to ground responses in site content. AskAny’s difference is what comes included in the free version: native WooCommerce order tracking by email or order ID, real-time SSE streaming that cuts server load compared to polling-based chat widgets, live agent handoff with full conversation history preserved on transition, and a choice of six AI providers, all without a Pro upgrade required to use any of it. It indexes your posts, pages, products, PDFs, and external sources first, then retrieves the right content for each question before the AI generates a response. Every answer includes a source citation, so visitors (and you) can see exactly where an answer came from. AskAny Website | Documentation | Pro Support AI Chatbot with RAG for WordPress Index – AskAny crawls your posts, pages, products, PDFs, Q&A pairs, and external URLs into a semantic search index automatically. Retrieve – When a visitor asks a question, AskAny finds the most relevant content chunks using hybrid semantic and keyword search. Respond – The AI generates a cited answer using only your indexed content. It cannot answer from general knowledge, so responses stay specific to your business. RAG is what separates an AI chatbot that answers accurately from one that hallucinates plausible-sounding nonsense. If a visitor asks about your return policy, a chatbot without RAG guesses. AskAny looks up your actual policy page first. Semantic Search and Content Indexing The RAG and semantic search system is what makes AskAny different from keyword-triggered chatbots. It understands meaning and context, so it can find the right answer even when a visitor’s question doesn’t exactly match the wording in your content. Content types indexed: posts, pages, custom post types, WooCommerce products and categories, custom fields, comments, menus, PDF documents, external URLs, Q&A pairs, and FAQs. Indexing is designed to keep API costs low. Content hash caching reuses embeddings whenever content has not changed, cutting API usage by up to 90%. Batch embedding processes multiple texts in a single API call. Smart update detection only re-embeds content that has actually been modified. Hybrid search runs both semantic understanding and keyword matching together. This covers cases where exact phrase matching is needed alongside conceptual retrieval, so a product SKU or an exact price still matches correctly. Live Agent Handoff for WordPress Chat Agents manage conversations from the WordPress admin chat panel. When a visitor needs a human, the agent takes over in one click. Full conversation history carries over instantly with no visible interruption for the visitor. The live agent system includes role-based access, online/offline status tracking, an unread message badge, and Slack integration so agents get notified and can reply without opening the WordPress admin. A Mobile App Push Bridge lets agents respond from Android. Unlike chatbot plugins that gate live agent seats or Slack notifications behind a paid tier, AskAny includes this in the free version. WooCommerce Order Tracking in Chat Customers can check their order status directly inside the chat widget. They can look up by email address, order ID, or transaction ID, with no need to visit a separate order tracking page. The WooCommerce integration also handles product information queries, real-time stock checks, smart product recommendations based on the conversation, and direct add-to-cart from chat. Product data is indexed automatically for RAG retrieval, so answers about pricing, stock, and specs stay current as your catalog changes. Telegram, WhatsApp, and AI-Assisted Live Chat AskAny now reaches visitors beyond the website widget, and gives live agents AI backup when they reply. Telegram Bot – Connect a Telegram bot under AskAny AI > Integrations, and visitors can message it directly. Every message routes through the same RAG pipeline as the website widget, so replies stay grounded in your indexed content, order data, and Q&A pairs instead of generic answers. WhatsApp via Twilio – Connect your Twilio WhatsApp number under AskAny AI > Integrations, and customers get the same AI-powered, cited responses over WhatsApp. No separate bot logic to maintain: WhatsApp, Telegram, and the website widget all draw from one content index and one conversation history. AI-Reviewed Compose for Live Agents – When a human agent takes over a conversation, they can draft a reply and run it through AI review before sending. The AI checks tone, clarity, and accuracy against the indexed content and suggests a refined version, so agents reply faster without sounding scripted. Attachments in Live Chat – Agents and visitors can now send file attachments directly inside the conversation, not just images. Attachments appear inline in both the admin chat panel and the visitor-facing widget, with previews available in conversation details. Six AI Providers, No Lock-In Switch between providers at any time without losing your indexed content or chat history. OpenAI – GPT-4o, GPT-4o-mini, GPT-4.1, O4-mini DeepSeek – Cost-effective with strong reasoning Google Gemini – Gemini 2.0 Flash, Gemini 2.5 Flash-Lite Anthropic Claude – Advanced reasoning, safety-focused X.AI Grok – Real-time knowledge OpenRouter – 100+ models through a single API Most competing WordPress chatbot plugins lock you into one AI provider or require the paid tier to switch. AskAny lets you change providers from Settings at any time with no re-indexing needed. Fallback Mode When your AI provider API is unavailable, AskAny keeps the chatbot running using indexed content and pattern matching. Visitors never see an error or a broken widget. Knowledge Base and Content Sources PDF Import – Upload PDFs and the plugin extracts and indexes every page automatically. Works for product manuals, documentation, policies, and specs. External Sources – Index any external URL, REST API endpoint, JSON feed, or XML source as an additional knowledge base. Q&A Management – Add custom question-answer pairs with Markdown support. Bulk import and export included. Automatic Site Indexing – WordPress content is indexed automatically on publish and on update. Analytics and Performance Track API request counts, success rates, token usage, and estimated costs per provider. Response time and popular question data helps you spot content gaps. Database tools include table size overview, one-click clean, optimize, and export, with date-range filtering. Customization Options The widget can be placed as a floating button (bottom-left or bottom-right) or embedded anywhere using the [askany_chat] shortcode. Configurable options include: custom avatar, logo, header text, welcome message, suggested messages with predefined Q&A responses, typing animation speed (Regular, Fast, Faster, Instant), and response tone (Human, Professional, Friendly, Robotic). A homepage FAQ section and custom links support Markdown. Full RTL and WCAG 2.1 AA accessibility compliance is included. Multi-Language Support The chatbot detects the visitor’s language and responds accordingly. Supported languages include English, Bengali, Spanish, French, German, Italian, Portuguese, Russian, Japanese, Chinese, Arabic, Hindi, and more. Competitor Analysis Search the WordPress plugin directory for an AI chatbot and a handful of names come up again and again, so it is worth being clear about where AskAny actually fits among them. AI Engine – The Chatbot, AI Framework & MCP for WordPress is the biggest name in the category by far, with chatbots, content generation, AI forms, and MCP server tools built on a huge add-on ecosystem. It is a genuine alternative to AskAny for chatbot use cases, but its RAG and vector database embeddings sit behind the Pro tier. AskAny includes RAG-based retrieval, WooCommerce order tracking, and live agent handoff in the free version, with a narrower focus on chatbot support rather than a full AI framework. MxChat – AI Chatbot & Content Generation for WordPress is the closest match. Its free version also trains on site content using RAG and includes WooCommerce integration and Slack-based live agent handoff, so anyone comparing the two is comparing like for like. Where AskAny pulls ahead is in the details: native order lookup by order ID, email, or transaction ID inside the chat widget itself, SSE streaming for instant responses instead of polling, and a Mobile App Push Bridge so agents can reply from Android without opening wp-admin. A newer wave of WordPress AI chatbots has also shown up in directory searches, and each falls into a different bucket relative to AskAny. AxiaChat AI, Hyve Lite, Bubblibot, Chatbot with ChatGPT WordPress (SmartSearchWP), and Vitxi Converse are all single-AI-provider chatbots, most locked to OpenAI, with WooCommerce order tracking, live agent handoff, and multi-provider support either absent or reserved for a paid tier. Hyve Lite and SmartSearchWP train only on posts and pages by default, with PDFs, custom post types, and WooCommerce sold as add-ons; Bubblibot ships a free core but keeps WooCommerce product indexing Pro-only; Vitxi Converse leans on templated FAQ branching rather than semantic RAG retrieval. AskAny includes RAG retrieval, WooCommerce order tracking, and six interchangeable AI providers in the free version, with none of that behind a paywall. AxiaChat AI, Muchat, and ILACHAT are the closest matches on the WooCommerce side, each combining an AI chatbot with product data and, in ILACHAT’s case, live human handoff and rich media messaging. None of the three currently offers native order lookup by order ID, email, and transaction ID together, SSE streaming, or a Telegram and WhatsApp bridge alongside the website widget, which is where AskAny’s free tier pulls ahead. WebChatAgent and Social Intents take a different architecture entirely: both are SaaS platforms with a WordPress plugin as the front door, so the chatbot logic, training data, and conversation history live on the vendor’s servers rather than in your own WordPress database. Social Intents in particular is built around routing chats into Slack, Microsoft Teams, Google Chat, or Zoom, with plans starting at $39/month once the trial ends. AskAny keeps everything, including chat history and the content index, in your own WordPress database, with no subscription required to use the AI chatbot, WooCommerce tracking, or live agent features. Beyond those two, most of the other names that show up nearby are solving a different problem entirely, not a weaker version of the same one. Wise Chat and Better Messages – Chat Rooms, Group Chat, Private Messages & AI Chat Bots are both built for people talking to each other, real-time chat rooms and private messaging between site members, with AI chat bots tucked away as a paid add-on rather than the core product. CleanTalk Anti-Spam. Spam Firewall & Bot Protection has nothing to do with chatbots at all, it blocks spam in comments, forms, and checkouts, and runs perfectly fine on the same site as AskAny. AI Agent by SiteGround is an admin-side tool, letting a site owner manage posts and WooCommerce products through a chat interface in wp-admin, not something a visitor ever sees. And AI Share & Summarize adds AI summaries and share buttons to blog posts, useful, but no chat widget involved. So of everything that turns up in an “AI chatbot for WordPress” search, only AI Engine and MxChat are actually built for the same job as AskAny: answering visitor questions from your own site content. Between those two, AskAny’s combination of free RAG retrieval, native WooCommerce order tracking, real-time SSE streaming, and live agent handoff, all in the free version, is what sets it apart. Privacy and Security All chat data is stored in your own WordPress database. Built-in GDPR controls cover data export, anonymization, and retention. Session security and XSS protection are included. See the External Services section below for exactly which third-party services are contacted, and only when you configure them. External Services This plugin connects to external services only for the features you configure. You supply your own API key for whichever AI provider you choose in Settings, and only that provider is contacted. OpenAI – Purpose: Generates chat responses and creates embeddings for semantic search when OpenAI is selected as the active AI provider. – Data sent: The visitor’s message, retrieved content snippets used for grounding, and site content submitted for embedding during indexing. – When it connects: On every chat response and during content indexing, only if OpenAI is the selected provider. – Service URL: https://api.openai.com/v1/ – Privacy Policy: https://openai.com/privacy/ – Terms of Service: https://openai.com/policies/row-terms-of-use/ DeepSeek – Purpose: Generates chat responses when DeepSeek is selected as the active AI provider. – Data sent: The visitor’s message and retrieved content snippets used for grounding. – When it connects: On every chat response, only if DeepSeek is the selected provider. – Service URL: https://api.deepseek.com/v1/ – Privacy Policy: https://www.deepseek.com/privacy-policy – Terms of Service: https://www.deepseek.com/terms-of-use Google Gemini – Purpose: Generates chat responses when Google Gemini is selected as the active AI provider. – Data sent: The visitor’s message and retrieved content snippets used for grounding. – When it connects: On every chat response, only if Gemini is the selected provider. – Service URL: https://generativelanguage.googleapis.com/ – Privacy Policy: https://policies.google.com/privacy – Terms of Service: https://ai.google.dev/gemini-api/terms Anthropic Claude – Purpose: Generates chat responses when Anthropic Claude is selected as the active AI provider. – Data sent: The visitor’s message and retrieved content snippets used for grounding. – When it connects: On every chat response, only if Claude is the selected provider. – Service URL: https://api.anthropic.com/ – Privacy Policy: https://www.anthropic.com/legal/privacy – Terms of Service: https://www.anthropic.com/legal/consumer-terms X.AI Grok – Purpose: Generates chat responses when X.AI Grok is selected as the active AI provider. – Data sent: The visitor’s message and retrieved content snippets used for grounding. – When it connects: On every chat response, only if Grok is the selected provider. – Service URL: https://api.x.ai/ – Privacy Policy: https://x.ai/legal/privacy-policy – Terms of Service: https://x.ai/legal/terms-of-service OpenRouter – Purpose: Routes chat responses to any of 100+ supported models when OpenRouter is selected as the active AI provider. – Data sent: The visitor’s message and retrieved content snippets used for grounding. – When it connects: On every chat response, only if OpenRouter is the selected provider. – Service URL: https://openrouter.ai/api/ – Privacy Policy: https://openrouter.ai/privacy – Terms of Service: https://openrouter.ai/terms Telegram – Purpose: Delivers AI-powered chat responses to visitors who message your connected Telegram bot. – Data sent: The visitor’s Telegram message, retrieved content snippets used for grounding, and the AI-generated reply routed back through the Telegram Bot API. – When it connects: When Telegram integration is enabled under AskAny AI > Integrations, on every message sent to your bot. – Service URL: https://api.telegram.org/ – Privacy Policy: https://telegram.org/privacy – Terms of Service: https://telegram.org/tos Twilio (WhatsApp) – Purpose: Delivers AI-powered chat responses to customers who message your connected Twilio WhatsApp number. – Data sent: The visitor’s WhatsApp message, retrieved content snippets used for grounding, and the AI-generated reply routed back through Twilio’s messaging API. – When it connects: When WhatsApp integration is enabled under AskAny AI > Integrations, on every message sent to your Twilio number. – Privacy Policy: https://www.twilio.com/en-us/legal/privacy – Terms of Service: https://www.twilio.com/en-us/legal/tos WPAzleen Settings API – Purpose: Loads display settings for the Pro upgrade modal in the plugin admin area. – Data sent: No personal …

The best AI Chatbot for WordPress. Like having ChatGPT trained on your content — turn your site into a 24/7 sales & support machine. 🤖 The best free AI Chatbot for WordPress If you want more leads, more sales, and less repetitive support work, you need a powerful chatbot on your website. AxiaChat AI is the best AI chatbot for WordPress designed to help you automate conversations, capture opportunities, and assist visitors instantly — 24 hours a day. This is not just another basic chatbot. This is a smart, business-ready chatbot built for real results — like having a private ChatGPT that knows your business inside out. ⚡ Why This Chatbot Is Different Most WordPress chatbot plugins only provide simple scripted replies. AxiaChat AI gives you a real AI-powered chatbot — backed by the same technology as ChatGPT — that understands context, learns from your content, and responds like a human assistant. Your chatbot can: ✅ Answer product questions automatically, 24/7 ✅ Guide customers to the right service or product ✅ Capture leads automatically — even when you’re offline ✅ Book appointments directly from the chat ✅ Analyze images & documents — visitors send photos or PDFs, the AI sees and responds ✅ Reduce support workload by handling repetitive questions ✅ Work 24/7 without breaks, staff costs, or delays This chatbot becomes part of your team. That’s what real AI chat for WordPress looks like. ⭐ What Users Say ⭐⭐⭐⭐⭐ “The only WooCommerce chatbot plugin that actually works” “Congratulations to the developer for an excellent job. The plugin is well built, reliable, and delivers exactly what it promises.” — gabricp100, February 2026 ⭐⭐⭐⭐⭐ “My conversion rate went up thanks to this chatbot!” “Perfect for small businesses. My conversion rate went up thanks to this chatbot!” — david25k, February 2026 ⭐⭐⭐⭐⭐ “Customizable and powerful” “I love how customizable it is, you can train your own AI agent and make it fit your brand.” — queroamilys, March 2026 ⭐⭐⭐⭐⭐ “It works great!” “We implemented it in the store and it works great.” — Antonio Aponte, February 2026 12 reviews on WordPress.org · 100% five-star ratings. Read all reviews → 📈 Built for Growth Whether you run an eCommerce store, agency, SaaS, or local business, this chatbot helps you scale without hiring more staff. Instead of answering the same questions every day, let your chatbot handle them. Instead of missing leads after hours, let your chatbot capture them. Instead of losing visitors, let your chatbot engage them instantly. A smart chatbot increases conversions. A fast chatbot improves user experience. A trained chatbot builds trust. 🎓 Train Your Chatbot on Your Own Content What makes this chatbot powerful is its ability to learn from your website. You can train your chatbot using: 📘 Pages and posts 🛒 WooCommerce products (prices, stock, SKU — always up to date) ❓ FAQs 📄 PDFs 🧠 Custom knowledge bases 🌐 External web pages And your visitors can send files directly in the chat: 📸 Photos & screenshots — the AI literally sees the image using vision AI 📄 PDF documents — text is extracted automatically and analyzed in context Imagine a customer sending a photo of their broken appliance, or uploading a contract for the AI to summarize. That’s not future tech — it works today. This means your chatbot gives accurate answers based on YOUR business — not generic internet responses. Your visitors talk to a specialized chatbot, not a random AI. Genuine AI chat powered by your own knowledge base. 🎯 Designed for Conversions This is not just a support chatbot. It is also a sales chatbot. Your chatbot can: 🎯 Collect names and emails 📧 Send notifications to your team 📅 Book appointments 👤 Escalate to a human agent � Receive photos & documents from visitors and act on them 🔗 Trigger custom workflows including forwarding uploaded images securely Every AI chat conversation with your chatbot becomes an opportunity. ⚡ Easy Setup in Minutes You don’t need technical skills to launch your chatbot. Install the plugin Add your API key Run the setup wizard Activate your chatbot In less than 5 minutes, your AI chatbot is live and your AI chat channel is open for visitors. 🎨 Fully Customizable Chatbot Match your brand perfectly. Your chatbot supports: 🎨 Custom colors and footer 🤖 Avatars (9 included + custom upload) � File upload button (images & PDFs) with in-chat preview �📱 Floating widget or inline embed via shortcode ✍️ Draggable, minimizable, and closable window 📲 Fully mobile responsive design Make your chatbot look like it was built specifically for your site. 💰 Control Costs & Usage Worried about AI costs? Your chatbot includes: 🔢 Daily usage limits (global and per-user) 🚫 Per-IP and per-user restrictions 📊 Real-time cost tracking and token usage 📝 Optional conversation logging (on/off anytime) You stay in control while your chatbot works for you. 🧠 Powerful AI Behind Your Chatbot Your chatbot can connect to: 🟢 OpenAI — GPT-5.6 Luna (default efficient chat), GPT-5.6 Terra/Sol, GPT-5.5, GPT-5.4 and legacy-compatible aliases 🟢 Anthropic Claude — Haiku 4.5 (default fast/cheap), Sonnet 5 (balanced), Opus 4.8 (premium), Fable 5 (flagship) and more 🟢 Google Gemini — 3.5 Flash (default fast chat), 3.1 Flash-Lite (lowest cost), 3.1 Pro and legacy-compatible 2.5 redirects You choose the AI engine for your chatbot. You control how your chatbot behaves. You decide what your chatbot knows. Unlike SaaS platforms or using ChatGPT directly, this chatbot runs directly on your WordPress site. Your data stays on your server. Your brand stays front and center. Your AI chat, your rules. 🔐 GDPR & Privacy Friendly This chatbot includes built-in consent tools and privacy controls. You can enable or disable conversation logging anytime. Your chatbot respects your users and your compliance requirements. ✅ Perfect For 🛒 WooCommerce stores 🏢 Agencies 🚀 SaaS businesses 🏤 Corporate websites 🔧 Service companies ✍️ Bloggers 🎓 Online educators If you want more engagement and automation, you need a professional chatbot and a reliable AI chat solution. 📚 Documentation & Technical Resources Official website: axiachat.org Full documentation, tutorials and advanced configuration guides: axiachat.org/help Community support forum: wordpress.org/support/plugin/axiachat-ai 🆓 Getting Started The free version of this chatbot is fully functional and covers most business needs. Start for free, explore all the core features, and grow at your own pace. 🚀 Stop Losing Opportunities Every visitor that leaves your site without interacting is a missed opportunity. A smart chatbot greets them. A helpful chatbot answers them. A proactive chatbot converts them. Install the best AI chatbot for WordPress today — your own ChatGPT working exclusively for your business — and transform your website into a 24/7 growth engine. Technical Features Supported AI Models OpenAI (GPT) GPT-5.6 Luna (default efficient chat) GPT-5.6 Terra GPT-5.6 Sol GPT-5.5 GPT-5.5 Pro GPT-5.4 GPT-5.4 Pro GPT-5.4 Mini GPT-5.4 Nano ChatGPT chat-latest GPT-5.2 Thinking Legacy aliases GPT-5.1/5.2/5.3 Chat Latest are auto-migrated to GPT-5.6 Luna GPT-5 GPT-5 Mini GPT-5 Nano GPT-4.1 / 4.1 Mini / 4.1 Nano GPT-4o / 4o Mini GPT-4 Turbo GPT-3.5 Turbo Anthropic (Claude) Claude Haiku 4.5 (default fast, lower cost) Claude Sonnet 5 (balanced chat and reasoning) Claude Sonnet 4.6 / 4.5 Claude Opus 4.8 (premium) Claude Fable 5 (flagship reasoning model) Claude Opus 4.7 / 4.6 Claude Opus 4.5 / 4.1 Claude Sonnet 4 and Claude Opus 4 — legacy (deprecated, retirement scheduled Jun 2026) Claude Sonnet 3.7 / Haiku 3.5 / 3 — legacy (retired) Google Gemini Gemini 3.5 Flash (stable, default fast chat) Gemini 3.1 Flash-Lite (stable, lowest cost fallback) Gemini 3.1 Pro Preview Gemini 3 Flash Preview — legacy (deprecated, use 3.5 Flash) Gemini 3.1 Flash-Lite Preview / Gemini 3 Pro Preview — legacy (shut down) Gemini 2.5 Pro / Flash / Flash-Lite — legacy (deprecated, retirement scheduled Oct 2026) Gemini 2.0 Flash / Flash-Lite — legacy (shut down Jun 2026) Automation & Integrations REST API ready — every bot interaction is driven by a WordPress AJAX endpoint, making it straightforward to integrate from external systems Outbound webhook parameters — configure method (GET/POST), headers, body template with dynamic placeholders, and response mapping Add-ons AI Tools — native function-calling tools (web search, calculations, date/time, etc.) File & Image Uploads — visitors can send photos (JPG, PNG, WebP) and PDF documents directly in the chat. Images are analyzed via AI vision; PDFs have text extracted automatically. Configurable per bot with file type and size controls Appointments — let the bot schedule appointments directly from the chat Lead Capture — collect and store visitor data (name, email, phone) inside the chat flow Web Scraper — fetch and summarise live web content as part of a bot response RAG — Retrieval-Augmented Generation Local vector search (cosine similarity) Pinecone support OpenAI embeddings Advanced indexing WooCommerce enrichment Featured images in chat Security & Validation WordPress nonce Honeypot protection CAPTCHA support Moderation API Rate limiting Usage limits Hidden system prompt Endpoint allowlist Frontend & UI Floating widget (configurable position: bottom-right, bottom-left, etc.) or inline layout via shortcode Fully draggable, minimizable and closable chat window Voice chat mode (type="voice_text") with browser Web Speech API File upload button (📎) with drag-and-drop support, in-chat thumbnail preview, and admin log attachment indicators Automatic image optimization (resize + compress) to minimize API costs while preserving visual quality Markdown rendering including images, bold, lists, and code blocks Suggestion chips / quick-reply buttons WhatsApp CTA button inside the chat window (configurable per bot) Admin & Developer Tools Guided Easy Setup Wizard — fully working bot in under 2 minutes One-click AI-generated bot instructions — describe your use case, AI writes the system prompt Built-in prompt templates (Customer Support, Lead Generation, E-commerce, FAQ, etc.) Chat conversation logs with full transcript, token usage, and cost breakdown Email alerts on new conversations — full transcript or summary, per-conversation or digest mode [aichat id="slug"] shortcode with 20+ data-attribute parameters (layout, position, colours, avatar, voice, maximized, etc.) Bundled Libraries Bootstrap (local) Bootstrap Icons (local) smalot/pdfparser (LGPLv3) for PDF extraction External Services / Data Disclosure This plugin can connect to the following third‑party APIs depending on which features you enable. You (the site owner) must supply the API keys. No keys are bundled and no traffic is proxied through a vendor server controlled by this plugin author. 1. OpenAI Used for: chat completions / responses, embeddings (context indexing), moderation (safety checks). Endpoints used (HTTPS): * https://api.openai.com/v1/chat/completions (legacy Chat Completions) * https://api.openai.com/v1/responses (Responses API – new unified endpoint if configured) * https://api.openai.com/v1/embeddings (document/post/PDF embedding + wizard indexing) * https://api.openai.com/v1/moderations (content moderation) Data Sent: * User prompt text (per message) and limited conversation history (trimmed for token control) * System / policy instructions (security + privacy policy + bot instructions) * Optional retrieved context snippets (only the selected top‑N chunks or page excerpt – never the full original document) * Embedding requests: raw chunk text produced from your site’s content or uploaded PDFs * Moderation: only the user prompt text (not the entire history) Data Retention (Your Server): * Conversation log rows (if logging enabled) including user prompt, model reply, timestamps, bot slug, session id, optional user id. IP (binary) only stored when per‑IP limits are turned on. Disable logging to stop storing new rows. * Embeddings table stores numeric vectors generated from your content (not reversible plaintext) plus the original chunk text for retrieval. Recommendations: * Update your privacy policy to disclose sending user prompts and limited site content to OpenAI for processing. * Disable logging or periodically purge if you process personal data. Legal / Docs: * Terms: https://openai.com/policies/terms-of-use * Privacy: https://openai.com/policies/privacy-policy * Usage Policies: https://openai.com/policies/usage-policies Opt‑Out / Control: * Remove the OpenAI API key in Settings to stop all OpenAI calls (bots will refuse to answer). * Disable conversation logging. * Limit context ingestion to non‑sensitive pages. 2. Anthropic (Claude) Used for: alternative chat completions via Claude models (messages API) when a bot provider is set to Anthropic/Claude. Endpoint used (HTTPS): * https://api.anthropic.com/v1/messages Data Sent: * A rewritten message array: system instructions + user prompt + condensed prior turns (trimmed) + optional retrieved context snippets. * Model identifier, max tokens / temperature style parameters. Headers: * x-api-key (your key) and anthropic-version (currently 2023-06-01 set in code). Retention (Your Server): * Same as OpenAI notes for conversation logging (the provider choice does not change local storage schema). Legal / Docs: * Terms: https://www.anthropic.com/legal/terms-of-service * Privacy: https://www.anthropic.com/legal/privacy * Usage Policy: https://www.anthropic.com/legal/aup Opt‑Out / Control: * Leave the Claude API key blank; those bots will fallback/refuse if provider requires it. * Switch provider per bot back to OpenAI. 3. Pinecone (Optional Remote Vector Store) Only used if you explicitly create a Context with remote type “Pinecone”. Local context mode (default) never contacts Pinecone. Endpoint Pattern: * Region/index specific HTTPS endpoints you enter (example placeholder: https://controller.pinecone.io and index query/upsert endpoints under *.pinecone.io). The plugin validates host against an allowlist containing pinecone.io (filter extendable). Used For: * Upserting embeddings (during context indexing / syncing) * Querying similar vectors when answering a question in that context. Data Sent: * Vectors (embedding arrays) and associated metadata (post/page IDs, titles, short chunk text) for upsert. * Query: the embedding vector of the user question + top‑K request. Retention: * Stored inside your Pinecone project (not on this plugin server). Local DB still keeps minimal reference metadata if you mix modes. Legal / Docs: * Terms: https://www.pinecone.io/terms/ * Privacy: https://www.pinecone.io/privacy/ * Security: https://www.pinecone.io/security/ Opt‑Out / Control: * Do not configure a remote Pinecone context; use local mode or page mode instead. * Delete the remote context to stop future upserts/queries. 4. Google Gemini (Optional AI Provider) Used for: alternative chat completions via Google Gemini models when a bot provider is set to Google/Gemini. Endpoint used (HTTPS): * https://generativelanguage.googleapis.com/v1beta/models/{model}:generateContent Data Sent: * System instructions + user prompt + condensed prior turns (trimmed) + optional retrieved context snippets. * Model identifier and generation configuration parameters. Headers: * API key passed as URL parameter (key=). Retention (Your Server): * Same as OpenAI notes for conversation logging. Legal / Docs: * Terms: https://ai.google.dev/gemini-api/terms * Privacy: https://policies.google.com/privacy * Usage Policy: https://ai.google.dev/gemini-api/terms Opt‑Out / Control: * Leave the Gemini API key blank; those bots will fallback/refuse if provider requires it. * Switch provider per bot to OpenAI or Claude. 5. Freemius SDK (Licensing & Analytics) This plugin uses the Freemius SDK for license activation, update delivery, and anonymous usage analytics (opt-in). Endpoints used (HTTPS): * https://api.freemius.com (license checks, updates, analytics) Data Sent (if opted in): * Site URL, WordPress version, PHP version, plugin version, and admin email. * No visitor/chat data is ever sent to Freemius. Legal / Docs: * Terms: https://freemius.com/terms/ * Privacy: https://freemius.com/privacy/ Opt‑Out / Control: * During activation, choose “Skip” to opt out of anonymous analytics. * License-related API calls are required only for premium features. 6. Google Sheets (Optional Lead Export) Only used if you enable the Google Sheets lead destination in the Leads add‑on. The plugin connects to Google APIs via OAuth 2.0 …