Most “free” AI chatbot tools eventually hit a wall: they require you to bring an OpenAI API key, pay per message, or cap you at 50 conversations per month before upselling. ConnectXeo's free plan is different — it runs on Ollama models hosted on our VPS, so you pay nothing for LLM inference.
What's Actually Free
On the free plan, you get:
- Access to Ollama-hosted open-source models (Mistral, Gemma, Qwen, and more)
- Full Shopify OAuth integration — live product, order, and customer data
- Web widget deployment with one line of JS
- Knowledge base with document uploads and web crawling
- 200 conversations per month
No OpenAI API key. No per-token billing. No hidden costs.
Step 1: Create Your Workspace
Sign up at ConnectXeo and create a workspace. A workspace is a container for your store — it holds your agents, integrations, and team members. Give it a name that matches your store.
Step 2: Create an Agent
Inside your workspace, create a new agent. Configure:
- Name: What the chat widget calls itself (e.g. “Emma from StyleShop”)
- Persona: A short description of the agent's role
- Tone: Professional, friendly, or casual
- Fallback message: What to say when the question is out of scope
Step 3: Connect Shopify
Navigate to Integrations → Shopify → Connect. You'll be redirected to Shopify's OAuth screen. Approve the permissions. ConnectXeo stores the access token (encrypted at rest) and your agent immediately gains access to live store data.
Step 4: Add a Knowledge Base
Go to Knowledge Base and add your store-specific content:
- Upload your returns/refund policy PDF
- Crawl your existing FAQ page
- Paste in your shipping policy text
This content is vectorized and stored — the agent will search it semantically when answering questions that aren't covered by live Shopify data.
Step 5: Deploy the Widget
Go to Deployments → New Deployment → Web Widget. Copy the one-line JavaScript snippet and paste it into your Shopify theme (Online Store → Themes → Edit Code → theme.liquid before </body>). The widget appears instantly.
Next Steps
Once deployed, monitor your agent's conversations in the dashboard. If you hit the free plan's limit or want to use GPT-4o for higher accuracy, upgrade to a paid plan and add your own API key — or let us manage it for you.