Turn chat questions into research reports for marketing and content teams. A user types a topic, and the system searches the web, reads pages, and writes a clear report you can use right away.

It starts with a chat message. An AI creates up to four precise search queries. The flow sends them to SerpAPI, formats the results, and fetches each page with Jina AI for clean text. Another AI extracts facts that match the topic, adds context from Wikipedia when useful, and uses short term memory to keep follow up questions on track. A final AI writes a structured summary with the key points and findings.

You need API keys for OpenRouter, SerpAPI, and Jina AI. Add the credentials in n8n, keep the default batching to control volume, and run a test chat to confirm the steps. Teams can use this for competitor briefs, content outlines, or market notes for sales. Expect faster turnarounds and less copy pasting, with hours of browsing cut down to minutes.

• Chat message trigger collects the user question and starts the run
• LLM via OpenRouter generates up to four precise search queries
• Code node parses the LLM output and prepares query chunks
• SerpAPI HTTP request searches Google and returns organic results
• Code node formats results and extracts the best URLs
• Split in Batches controls loop volume for SerpAPI and Jina AI calls
• Jina AI request fetches clean page text for faster analysis
• LLM agent extracts only content that matches the user goal
• Wikipedia tool adds helpful background context when needed
• Memory buffers keep session context for better follow up answers
• Final LLM writes a structured research report ready to share

• Reduce manual research from 3 hours to 10 minutes
• Automate up to 80 percent of web reading and note taking
• Improve relevance by 30 percent using guided search and extraction
• Handle 10 times more topics per day with batching
• Connect SerpAPI, Jina AI, Wikipedia, and OpenRouter in one flow

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with OpenRouter, SerpAPI, Jina AI and Wikipedia. See the Tools Required section above for links to create accounts with these services.
3. Open the LLM Response Provider node. In the Credential to connect with dropdown, click Create new credential for OpenRouter. Paste your OpenRouter API key from your OpenRouter settings. Save and test the connection.
4. Set up SerpAPI. In the Perform SerpAPI Search Request node, ensure an API Key credential named SerpAPI exists. If not, create a new API Key credential in n8n with the key from your SerpAPI dashboard. Keep engine set to google and adjust location or language if needed.
5. Configure Jina AI. Open the Perform Jina AI Analysis Request node. In Credential to connect with choose HTTP Header Auth and create a new credential. Set Header Name to Authorization and Value to Bearer YOUR_JINA_API_KEY. Save and test.
6. Wikipedia needs no credential. The Fetch Wikipedia Information tool works out of the box. Keep it connected to the report step.
7. Check the Generate Search Queries using LLM node. Make sure it reads the user message from the Chat Message Trigger and limits to four queries.
8. Verify the Parse and Chunk JSON Data and Format SerpAPI Organic Results code nodes. Confirm they output query chunks and a clean list of URLs.
9. Review both Split In Batches nodes. Start with small batch sizes to avoid rate limits. Increase later if you need more throughput.
10. Open the memory buffer nodes. Keep the session key the same across both memory nodes so the chat keeps context.
11. Test the trigger. Open the Chat Message Trigger, start listening, and send a sample topic. Watch the run. You should see SerpAPI results, Jina AI page text, extracted context, and a final report.
12. Troubleshoot common issues. If no search results, check the SerpAPI key and query text. If Jina AI fails, confirm the Authorization header. If the report is thin, raise batch size or adjust the prompts in the extraction and report nodes.
13. Activate the workflow when tests pass. Share the chat entry point with your team for daily research requests.