# Chat4Agent — Chat Widget for Make, n8n & Zapier AI Agents > Turn your Make.com, n8n, Zapier, Activepieces or Gumloop AI agent into a branded chat widget. Paste a webhook, customise, embed. Free, no code, GDPR-ready. Source: https://chat4agent.com/ --- # Your AI agent, # finally has a face. You built the brain on Make.com, n8n or Zapier. Chat4Agent gives it a chat window your visitors can actually talk to: branded, embeddable, live in minutes. No code. [Start free](/signup)[See how it works](#how-it-works) - Free forever - No credit card - GDPR-ready POST https://hook.eu1.make.com/a8f3… The live widget: intro messages, a visitor question and the typed answer Connected ## Plugs into the tools you already trust - Make.com - n8n - Zapier - Activepieces - Gumloop or any MCP-capable agentic platform ## From webhook URL to live chat in three steps. No backend to build, no SDK to learn. The agent stays where it is; only its face is new. ### 1 ### Connect your agent 1. Paste the webhook URL of the scenario you already run. Chat4Agent sends a test message and tells you the moment the reply comes back well-formed. 1. Pasting an agent webhook URL and running Test Connection until it reads Valid ### 2 ### Make it look like you 2. Pick the font, the colours, the header and the launcher. The preview on the right follows every change, so what you see is what your visitors get. 2. Changing the primary colour and font while the live preview follows ### 3 ### Put it on your site 3. Copy one snippet for a floating button, an inline box or a full page. Or share the hosted page and skip the embedding altogether. And you are live. 3. Copying the embed snippet and the widget appearing on a page ### See how it works with Make.com One question, nothing cut: the visitor asks, the scenario runs module by module, the answer is back in the widget. Real time, side by side: a visitor asks in the widget, the Make.com scenario runs, the answer comes back ### See how it works with n8n The same conversation on a self-hosted n8n: the webhook fires, the AI Agent node runs, the answer lands in the chat. Real time, side by side: a visitor asks in the widget, the n8n workflow runs, the answer comes back ## Everything a chat needs. Nothing you have to build. ### Make it yours Fonts, colours, header image, launcher style and the wording of every label. Text colours are recomputed for contrast when you change a background, so it stays readable without you checking. Style & Velocity: font, both background colours, the launcher, the loader and the opening animation, with the preview following every change ### Say hello first Intro messages greet the visitor before they type, with quick-reply buttons for the questions you get most. Typing speed, bubble timing and twenty opening animations are yours to set. Copies: the header, the greeting and the wording inside the chat, then the intro messages replaying with their delays ### Put it anywhere A floating button in the corner, a box inline with your content, a full page, or a hosted page at chat4agent.com/p/your-name that needs no website at all. Embed & Install: the four ways to install, the snippet, private mode and the consent banner ### See what happens Every conversation and every event, searchable, with the timing of each reply. Send the events to your own webhook when you want them in your tools too. Conversations: the activity chart — ranges, tooltip and legend — then the log of every event underneath ## Built to be trusted with your visitors' conversations. The chat sits on your site under your name. Here is what keeps it, and the people using it, safe. ### Signed requests - Every call to your agent carries a secret only you hold. Rotate it whenever you like. ### Allowed origins - Your widget answers only from the domains you list. Anywhere else, it stays silent. ### Private chats - Lock a chat behind an access password for a team, a client or a course. ### Consent first - A GDPR consent notice in your own words before the first message is sent. ### Hosted in the EU - Your configuration, media and logs live on servers in Amsterdam. ### No card data kept - Payments run through Stripe. We store the card brand and its last four digits, nothing more. ### Retention you set - Keep logs for 24 hours, 90 days or nothing at all. The choice is yours, per organization. ### Take it with you - Export your full configuration and your logs any time. Delete a chat and its data goes with it. ### And you decide who gets in Name the domains the widget may run on, put a chat behind a password, and hand that password out yourself. Visibility & Security: allowed origins, private mode and the access password it generates ## Any use case, any face. The same platform, three very different chats. - Floating button - ![A store support chat in green, opened from the floating button on the shop’s own site](https://chat4agent.ams3.digitaloceanspaces.com/system/landing/Chat4Agent.com_usecase-support.png) - A store support chat in green, opened from the floating button on the shop’s own site ### Order help for a store - A Make.com scenario reads the order, the chat answers in the brand colours from the shop's corner. - Make.com - Dedicated page - ![An internal HR assistant in blue, on a hosted page kept private behind an access password](https://chat4agent.ams3.digitaloceanspaces.com/system/landing/Chat4Agent.com_usecase-hr.png) - An internal HR assistant in blue, on a hosted page kept private behind an access password ### An HR assistant on the intranet - An n8n workflow over the policy handbook, private behind a password, on a hosted page. - n8n - Inline in the page - ![A lead qualifier in rose, embedded inline in a landing page rather than floating over it](https://chat4agent.ams3.digitaloceanspaces.com/system/landing/Chat4Agent.com_usecase-leads.png) - A lead qualifier in rose, embedded inline in a landing page rather than floating over it ### A lead qualifier on a landing page - A Zapier zap asks three questions inline and drops the answers into the CRM. - Zapier ## Free forever. Pay only for polish and power. Unlimited widgets and conversations on Free, under fair use. Prices excl. VAT. Monthly Yearly, 2 months free ### Free **€0**forever A complete chat for one agent. Not a trial. [Start free](/signup) - Unlimited chat widgets - Unlimited conversations - Dedicated chatbot page - Floating button widget - Inline embed - Full-page embed - Webhook connection to AI agent [Compare every feature on the pricing page](/pricing) ## Questions people ask before signing up. Is the free plan really free? Yes. No credit card and no time limit. Free includes unlimited chat widgets and unlimited conversations under fair use. You pay only when you want Pro or Expert features such as removing the branding or private chats. Do I need to write code? No. You paste the webhook URL of the agent you already built, adjust the look in the editor with a live preview, and copy one snippet onto your site. A hosted page is ready without any embedding at all. Which tools does it work with? Make.com, n8n, Zapier, Activepieces, Gumloop and anything else that can receive an HTTP webhook. The Expert plan also connects to MCP-capable agentic platforms, OpenAI agents and Mistral agents. Who can read the conversations? Messages travel from the widget to your agent signed with a secret only you hold, and your widget only answers from the domains you allow. Logs are kept for the period you set, from 24 hours to 90 days, and you can export or delete them. Can I use my own domain? Yes, on the Expert plan. Dedicated chat pages can be served from your own domain instead of chat4agent.com. How do I cancel? From Billing, in one click. Paid features stay on until the end of the period you already paid, then the organization returns to Free with all its chats intact. ## Your agent is ready. Give it a face. Five minutes from webhook to a chat your visitors will actually use. [Start free](/signup)