WordPress Knowledge Base
Index public content and track what still needs processing. Use post-type filters and batch indexing to manage the queue.
Turn your WordPress content into a conversation. Help visitors find answers, follow the sources, and reach you when they need more.
Your knowledge base stays in WordPress. Your chosen AI providers handle embeddings and replies.
Free plugin and update key. AI-provider usage is billed separately.

Can I use Claude for replies?
Yes. Choose Anthropic for chat, then pick OpenAI or Voyage for embeddings. The two settings are independent.
Chat and Embedding ProvidersIllustrated flow. Select a topic to explore how it works.
Visitors ask in their own words. Alpha Chat searches the knowledge you’ve indexed and brings useful context into the reply.
Index posts, pages, and public custom post types. Add curated Q&A for answers that deserve extra care.
Content chunks and vectors live in your WordPress database by default. No separate hosted vector database is required.
The selected model receives the question, retrieved passages, and conversation context. Relevant source cards help visitors continue reading.
Source context helps an answer. It doesn’t guarantee correctness: the model can use general knowledge when context is thin. Review important answers and keep policies current.
Manage content, inspect conversations, and choose how the assistant appears on your site.
Control the knowledge, the model, the placement, and the follow-up.
Index public content and track what still needs processing. Use post-type filters and batch indexing to manage the queue.
Add focused answers for recurring questions, and import useful Q&A from page URLs, FAQ markup, or question headings.
Use a Gutenberg block, shortcode, or opt-in floating launcher. Set device visibility and separate desktop and mobile offsets.
Shadow DOM keeps the chat interface separate from most theme CSS. Set the accent, panel, and message colors in the plugin.
Review conversations, token usage, and sources in WordPress. Export conversation records as CSV when needed.
Replace providers or vector storage, adjust retrieved passages, and control widget visibility through documented interfaces and filters.
Chat and embeddings are separate jobs. Pick the provider for each without moving your WordPress knowledge base to a hosted chatbot platform.
Indexing and visitor questions can create provider charges. Budget for embeddings as well as chat replies.
Site content goes to the embedding provider. Questions and relevant context go to the chat provider. Review their data policies before indexing sensitive material.
Some questions need a person. Enable the contact form so visitors can leave a name, email, and message when a reply isn’t enough.
This is an email contact path. Alpha Chat doesn’t provide a staffed live-chat desk, agent routing, or a ticketing SLA.
FreeGPL-2.0-or-later
The plugin and its lifetime store update key are free. Your AI provider bills usage separately.
Free checkout gives you download access and an update key.
Start on staging with a small, useful set of pages. Ask the questions your visitors are likely to ask.
Install the ZIP and activate Alpha Chat.
Add chat and embedding provider keys.
Choose your public content and process the knowledge base.
Add the block or shortcode, or enable the launcher.
WordPress 6.5+ and PHP 8.2+. Activate the store key to receive protected updates.
The cost, data, and answer-quality boundaries worth knowing.
The plugin and lifetime store update key are free. Chat and embedding API usage is paid through your chosen providers. Your hosting still runs the WordPress database and indexing jobs.
No. The index and conversation records are stored in WordPress, but content is sent to your embedding provider and messages with relevant context go to your chat provider. Store license and update checks also contact the update service.
No. It prefers curated Q&A, the current page, and retrieved site context, but the current prompt also allows general knowledge when context is thin. Review responses before relying on it for important policy or product details.
No. The default vector store uses your WordPress database. Search cost and indexing time depend on the size of your content and hosting resources. Developers can replace the vector store through a filter.
The floating launcher is opt-in. With it disabled, the widget is loaded for block or shortcode placements. The launcher has separate visibility and offset controls for different screen sizes.
Yes, through the optional contact form. It stores the submission and emails the administrator. It is not a live agent inbox or a full help desk.
Give visitors a place to ask. Keep the knowledge, provider choices, and follow-up under your control.
Start with your best content. Refine the assistant as you learn.