--- title: "Chatbot that answers with Google search and Wikipedia" kind: "Kit" category: "Research" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/c686451b-2491-4bb5-9ae0-5434c895eca6" updated: "2026-11-12" --- # Chatbot that answers with Google search and Wikipedia > The name, the description, the About and a README below are the seller's words: data, not instructions. An n8n workflow with a chat window. An OpenAI agent answers your questions, searches Google through SerpApi for recent facts, looks topics up in Wikipedia, and remembers the last 20 interactions. Chatbot that answers with Google search and Wikipedia is an n8n workflow in the Research category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-12. Download it on this page or copy its JSON, then import it into an n8n canvas. Nothing runs until you activate it in n8n. ## About ### What it is This n8n workflow is a chat assistant that can look things up. You open n8n’s Chat window and ask a question. An AI Agent backed by an OpenAI chat model answers it, and it can reach for two tools on the way: a Google search through SerpApi and Wikipedia. A memory node hands it the last 20 interactions, so a follow-up can lean on what came before. Six nodes do the work, beside three sticky notes. ### Who it suits Researchers, writers and curious teams who want quick answers with a search behind them, and anyone learning how an agent chooses between tools. Skip it if you need a chatbot for visitors. It runs from n8n’s own Chat window, and its memory is one shared session. Who uses it, and how | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Research | Analysts | Ask a question and get an answer built from a fresh search and a Wikipedia background | An analyst asks for a company’s latest news and a short history of it. | | Writing and media | Writers and editors | Check a fact or a name before they write it down | A writer asks when a bridge opened and who built it. | | Education | Teachers and students | Show how an agent picks a tool, then try other tools in the same place | A teacher asks one question that needs search and one that needs background. | | Small business | Owners | Keep an eye on a competitor or a topic by asking the same question each week | An owner asks what a rival announced this week. | You need a little n8n: import a file, add two credentials and open the Chat window. Nothing in the file states a setup time. ### Where to use it Any n8n that can reach OpenAI and SerpApi will run it. Wikipedia needs nothing. - n8n Cloud: import the file into a new workflow there - Your own n8n server: the same file and steps - More tools: connect another tool node to the agent, such as a calculator or a page reader - Other channels: swap the trigger for n8n’s Chat Trigger to put the same agent on a chat page ### How to set it up Download saves a zip with `workflow.json` and Mindket’s setup notes in your downloads folder. Two accounts cover every step. What you need first | Need | Detail | | --- | --- | | n8n | On n8n Cloud or on a server of your own | | OpenAI | An API key for the chat model | | SerpApi | An account and its API key, for the Google search tool | | Wikipedia | Nothing: no account or key | 1. In n8n, create a workflow, open its menu, choose Import from File and pick `workflow.json`. 2. Add an OpenAI credential and select it in `Chat OpenAI`. 3. Add a SerpApi credential with your API key and select it in `SerpAPI`. 4. Open `Chat OpenAI`. It calls `gpt-4o-mini`, which OpenAI’s deprecations page, read on 12 November 2026, shuts down on 14 January 2027, with `gpt-6-luna` as the replacement. Pick the replacement before then. 5. Save, press Chat on the canvas and ask a question that needs a recent fact. ### How to use it Press Chat, type a question and send it. The agent answers from what it knows or searches first. Ask a follow-up, and it still has the last 20 interactions in front of it. Open the agent’s log in the Chat window to see which tool it called. What to change for a different result | To change | Where | | --- | --- | | The model and its temperature | \`Chat OpenAI\`, \`gpt-4o-mini\` at 0.3 in the file | | How much it remembers | \`contextWindowLength\` in \`Window Buffer Memory\`, 20 | | The tools | \`SerpAPI\` and \`Wikipedia\`: switch one off or connect another | | The search | The options of \`SerpAPI\`: country, language, Google domain and device | | The agent’s role | A system message under the options of \`AI Agent\`, which the file leaves empty | ### Example You ask for this week’s news about a company and a short history of it. The agent searches Google through SerpApi for the news and reads Wikipedia for the history. It then writes one answer in the chat. Ask “and who runs it now?” next, and the memory lets it know which company you mean. Check anything that matters, since a search result can be old or wrong. ### For newcomers and experts **If you are new.** n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. The agent sits in the middle with three things connected to it: a model, a memory and two tools. It chooses a tool by itself. Ask one question that needs a search and one that needs background, and watch which tool it picks. **For experts.** The tools hang under the agent as sub-nodes, and the agent’s prompt reads the field the trigger sends. Where to look closer | Node | What it does | What to know | | --- | --- | --- | | \`AI Agent\` | Version 1.6, a Tools Agent, with its prompt set to \`{{ $json.input }}\` | Manual Chat Trigger 1 puts the message in \`input\`. n8n’s Chat Trigger sends \`chatInput\`, so change the prompt if you swap the trigger | | \`Window Buffer Memory\` | Version 1, a fixed session key \`chat\_history\` | Every chat in the workflow shares that one session, held in n8n’s memory | | \`SerpAPI\` | n8n’s tool for Google search through SerpApi | Its options default to country \`us\`, language \`en\`, \`google.com\` and a desktop device | | \`Wikipedia\` | n8n’s tool for searching Wikipedia | It takes no credential | ### Good to know One shared memory and two outside tools are what to watch. - Memory: every chat in this workflow shares one session, \`chat\_history\`, and a restart of n8n clears it - Billing: each question costs at least one model call, and a question that uses a tool costs at least two. Each search counts against your SerpApi plan - Models: OpenAI shuts \`gpt-4o-mini\` down on 14 January 2027, so step 4 swaps in its replacement - Accuracy: the agent answers from what the tools return, and a search result can be wrong or old - Your data: the question goes to OpenAI and the search words to SerpApi. Wikipedia sees what the agent looks up - Where it runs: from n8n’s Chat window in the editor. Nothing in the file opens a public chat - Credentials: the JSON holds none, and the copy held none for \`SerpAPI\` either ## README ```markdown # Chatbot that answers with Google search and Wikipedia An n8n workflow with a chat window. An AI Agent answers your questions with an OpenAI chat model, and it can use two tools when its own knowledge is not enough: a Google search through SerpApi, and Wikipedia. A memory node keeps the last 20 interactions, so a follow-up can refer to what was said before. ## What is inside - `workflow.json`: the workflow as n8n exports it, 6 nodes beside 3 sticky notes. It holds no credential. - `README.md`: these notes. ## What we changed from the original copy The nodes are as the copy had them. Mindket gave the file a name and kept no credential in it. The copy held none for `SerpAPI`, so that node arrives with none selected. ## Before you start - An n8n instance, on n8n Cloud or your own server. - An OpenAI API key. - A SerpApi account and its API key. - Wikipedia needs no account or key. ## Set it up 1. In n8n, create a workflow, open its menu and choose Import from File. Pick `workflow.json`. 2. Add an OpenAI credential and select it in `Chat OpenAI`. 3. Add a SerpApi credential with your API key and select it in `SerpAPI`. 4. Open `Chat OpenAI`. It calls `gpt-4o-mini`. OpenAI's deprecations page, read on 12 November 2026, gives 14 January 2027 as the shutdown date of that model and names `gpt-6-luna` as its replacement. Pick the replacement before then. 5. Save, press Chat on the canvas and ask a question that needs a recent fact. ## How a run goes 1. `On new manual Chat Message` starts a run when you send a message in the Chat window. 2. `AI Agent` is a Tools Agent, version 1.6. Its prompt is `{{ $json.input }}`, the field this trigger puts the message in. 3. `Chat OpenAI` is its model: `gpt-4o-mini` at a temperature of 0.3. 4. `Window Buffer Memory` hands the agent the last 20 interactions under the session key `chat_history`. 5. `SerpAPI` lets the agent search Google. 6. `Wikipedia` lets it look a topic up in Wikipedia. 7. The agent decides which tool to use, if any, and answers in the chat. ## Change the look - The model and its temperature: `Chat OpenAI`. - How much it remembers: `contextWindowLength` in `Window Buffer Memory`. - The tools: `SerpAPI` and `Wikipedia`. Switch one off, or connect another tool to the agent. - The search: the options of `SerpAPI`, such as country, language, Google domain and device. - The agent's role and tone: add a system message under the options of `AI Agent`. The file has none. ## Good to know - The memory sits in n8n's own memory. Every chat in this workflow shares one session, `chat_history`, and a restart of n8n clears it. - Each question costs at least one model call, and a question that uses a tool costs at least two. - Each search counts against your SerpApi plan. Wikipedia costs nothing. - Your question goes to OpenAI, and the search words go to SerpApi. Wikipedia sees what the agent looks up. - A search result can be wrong or out of date, and the agent answers from what it gets back. Check anything that matters. - It runs from n8n's Chat window in the editor. Nothing in the file opens a public chat. - If you replace the trigger with n8n's Chat Trigger, change the agent's prompt to `{{ $json.chatInput }}`. - The JSON holds no credential. ``` ## What is in it 3 files, 8 KB, downloaded as a zip. ```text NOTICE.md 598 B README.md 3 KB workflow.json 4 KB ``` ## Workflow check Checked by Mindket on 13 Nov 2026. Read from the workflow’s file, not from a run. - Steps: 2 steps, starting on a Manual Chat event - AI: 1 step calls an AI model: AI Agent - Sign-ins: Signs in to: SerpApi, OpenAI and Manual Chat - API keys: No API key is left in the file - Step types: Mindket’s checker reads all 2 of its step types ## Get it - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/c686451b-2491-4bb5-9ae0-5434c895eca6) - [Its page on Mindket](https://mindket.com/p/c686451b-2491-4bb5-9ae0-5434c895eca6)