Build a Free AI Customer Support Bot — No-Code, Self-Hosted, $0 in 2026

This post is part of AIGetFree's ongoing coverage. Browse our Blog or check the API Gateway Showdown.

The short version: Everything you need to know, tested and verified.

This guide is part of our $0 Stack series for building useful AI systems without monthly software bills. Read the companion guides: The $0 AI Automation Stack, The Ultimate AI Agent Stack, and Build an AI Agent With n8n.

Go to cloud.dify.ai and sign up with your email address. You will land on the Studio workspace — a clean dashboard showing your apps, knowledge bases, and tools. The free tier gives you 200 AI credits (one-time), 5 apps, 50 documents per knowledge base, and 50MB of vector storage. For a support bot handling hundreds of questions per day, this is more than enough — the LLM calls go through OpenRouter, not Dify credits.

Self-hosted alternative: If you want unlimited everything and complete data control, deploy Dify with Docker Compose. Clone the repo from github.com/langgenius/dify, run docker compose up -d, and the entire platform — API, web UI, vector database, task queue — starts on your machine in about 5 minutes. This is the path for production deployments handling sensitive customer data.

This is the step that makes the entire stack free. In Dify, click your avatar (top-right) → Settings → Model Provider in the left sidebar. Click Go to Marketplace, search for “OpenRouter”, and click Install. The plugin installs in seconds. Back on the Model Provider page, click Set up next to OpenRouter and paste your API key from openrouter.ai/keys.

Now the critical configuration. Scroll to System Model Settings at the top of the Model Provider page. Set System Reasoning Model to google/gemini-2.0-flash-001 (free on OpenRouter, excellent for customer support). Set Embedding Model to openai/text-embedding-3-small — this costs approximately $0.02 per million tokens, meaning your entire knowledge base embedding will cost less than one cent. If you want truly zero cost, use a free embedding model from OpenRouter like nomic-ai/nomic-embed-text-v1.5.

Click Knowledge in the top navigation bar, then Create Knowledge. This is where you upload the documents that ground your bot in reality. Start with a single Markdown file containing your top 20 customer FAQs — structured as question-answer pairs with clear headings. Dify will chunk this into searchable segments automatically.

Supported formats: .txt, .pdf, .md, .html, .docx, .csv. Upload your help center articles, refund policy, shipping information, product specs, troubleshooting guides, and onboarding documentation. On the Chunk Settings page, leave the defaults: High Quality indexing mode (uses your embedding model for semantic search), 500-token chunks with 50-token overlap. For retrieval, select Hybrid Search — this combines semantic similarity (“what does this question mean?”) with keyword matching (“does this document contain the exact term ‘password reset’?”), which is essential for support bots where customers may phrase things differently than your docs.

After uploading, Dify processes the documents — chunking, embedding, and indexing. For 20 FAQs, this takes about 30 seconds. You will see a confirmation when the knowledge base is ready. Test it immediately: click “Test Retrieval” and type a question. Dify shows you which chunks it retrieved and how relevant they are. If the right chunks appear, your knowledge base is working.

Go to Studio → Create App → select Chatflow. This opens a visual canvas where you wire nodes together. Chatflow is the right choice (not Chatbot) because it supports Knowledge Retrieval nodes and conditional branching — both essential for a support bot that cites sources and escalates to humans.

Build the node chain in this exact order:

Node 1: Start — The entry point. Configure the opening message: “Hi! I am the [Company Name] support assistant. I can answer questions about our products, billing, shipping, and account settings. What can I help you with?” Add suggested questions below: “How do I reset my password?”, “What is your refund policy?”, “How long does shipping take?”

Node 2: Knowledge Retrieval — Connect this to Start. Select your knowledge base from the dropdown. Set Top K = 3 (retrieve the 3 most relevant chunks). Enable Rerank if available — this re-scores retrieved chunks for better relevance. Set Score Threshold = 0.5 — chunks below this relevance score are discarded, preventing the bot from using irrelevant passages.

Node 3: LLM — Connect Knowledge Retrieval output to LLM input. This is where you write the system prompt that defines your bot’s behavior. Use this exact template:

Node 4: Answer — Connect the LLM output to Answer. This is what the customer sees. Configure the output to show the LLM response directly.

This is the step that separates a useful bot from a dangerous one. Without handoff logic, your bot will hallucinate answers when it does not know something — exactly what you are trying to prevent. Dify Chatflow supports conditional branching through the IF/ELSE node.

Insert an IF/ELSE node between the LLM and Answer nodes. Configure the condition: IF LLM output contains "human agent". On the TRUE branch: add a Webhook node that sends a POST request to your Slack/Discord/Email webhook with the customer’s question and conversation history. Then add an Answer node that tells the customer: “A member of our support team will respond within [timeframe]. Your reference number is [conversation ID].” On the FALSE branch: route directly to the normal Answer node.

Slack webhook setup: Go to Slack → Apps → Incoming Webhooks → Add to Workspace → copy the webhook URL. Paste it into the Dify Webhook node. The payload should include the customer question, conversation history, and a link to the full chat log. Your support team gets a notification within seconds of the escalation.

Click Publish in the top-right corner of your Chatflow. Dify generates everything you need: a shareable URL for testing, a JavaScript snippet for embedding, and a REST API endpoint for custom integrations. Click the Embed tab and copy the JavaScript code.

WordPress: Paste the snippet into your theme’s footer (Appearance → Theme File Editor → footer.php) or use a Custom HTML block in your page builder. Webflow: Add an Embed element to your page and paste the code. Shopify: Go to Online Store → Themes → Edit Code → theme.liquid, paste before the closing </body> tag. Custom HTML: Paste anywhere in your HTML file.

Customize the widget appearance before deploying: set the bubble color to match your brand, change the greeting message, adjust the position (bottom-right is standard), and upload your company logo. The widget is responsive — it works on mobile, tablet, and desktop. Test it on your staging site first, then deploy to production.

The Golden Rule of Support Bots

Start with 20 FAQs, test for one week, then expand the knowledge base based on real customer questions — not what you think customers ask. The best support bot grows from the language customers actually use, not the language you put in your documentation. Review conversation logs weekly and add every question the bot could not answer.

The Bottom Line

In 20 minutes and $0, you can deploy a RAG-powered customer support bot that answers from your own documentation, cites sources, and escalates to humans when needed. This is not a demo — it is production-ready for small businesses, SaaS startups, and internal teams. The stack is Dify (visual builder) + OpenRouter (free LLM) + your existing documentation. Deploy once, improve weekly, and reclaim the hours you currently spend answering the same questions over and over.

Try Dify Free →