{"product_id":"n8n-workflow-multi-llm-debate-validate-chat-answers","title":"n8n Workflow: Multi-LLM Debate \u0026 Validate Chat Answers","description":"\u003ch3\u003eGet chat answers that survive a structured debate—automatically\u003c\/h3\u003e\n\u003cp\u003eThis n8n workflow runs a multi-LLM “debate \u0026amp; validate” process: an OpenAI (GPT) model drafts an answer, AWS Bedrock Claude Sonnet challenges it, and AWS Bedrock Claude Opus makes the final approve\/reject decision—returning either the approved answer or a corrected replacement.\u003c\/p\u003e\n\n\u003ch3\u003eWhat this workflow does\u003c\/h3\u003e\n\u003col\u003e\n  \u003cli\u003e\n\u003cb\u003eReceives a chat message\u003c\/b\u003e via an n8n chat trigger, including the user’s question and optional settings like \u003cb\u003edomain\u003c\/b\u003e, \u003cb\u003emaxRounds\u003c\/b\u003e, and \u003cb\u003emaxCycles\u003c\/b\u003e.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eDrafts the first answer with OpenAI (GPT)\u003c\/b\u003e, optionally incorporating feedback from previous debate rounds.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eEvaluates the draft with AWS Bedrock Claude Sonnet\u003c\/b\u003e, which returns a structured verdict containing agreement status and objections.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eRuns debate loops (up to maxRounds)\u003c\/b\u003e: if Sonnet disagrees and rounds remain, the workflow feeds Sonnet’s objections back into GPT and repeats the answer-and-evaluate cycle.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eFinal decision with AWS Bedrock Claude Opus\u003c\/b\u003e: once Sonnet agrees (or rounds are exhausted), Opus decides to approve or reject, and produces a final answer.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eOptional correction cycles (up to maxCycles)\u003c\/b\u003e: if Opus rejects and cycles remain, its reason is passed back to GPT and the debate\/judging loop repeats until an approved (or corrected) response is returned with a status note.\u003c\/li\u003e\n\u003c\/ol\u003e\n\n\u003ch3\u003eUse cases\u003c\/h3\u003e\n\u003cul\u003e\n  \u003cli\u003e\n\u003cb\u003eQA for customer support\u003c\/b\u003e: validate drafted responses to reduce misleading or low-quality answers.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eKnowledge-heavy chatbots\u003c\/b\u003e: ensure answers are checked against critical feedback before replying to users.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eInternal research assistants\u003c\/b\u003e: improve reliability by forcing a “challenge then rule” process across multiple LLMs.\u003c\/li\u003e\n\u003c\/ul\u003e\n\n\u003ch3\u003eTechnical details\u003c\/h3\u003e\n\u003cul\u003e\n  \u003cli\u003e\n\u003cb\u003en8n Chat Trigger\u003c\/b\u003e to receive the question and routing parameters.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eOpenAI (GPT) via n8nn8n-nodes-langchainlm chat open ai\u003c\/b\u003e for the initial draft (and revised drafts after objections).\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eAWS Bedrock Claude Sonnet\u003c\/b\u003e for structured critical evaluation and objections.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cb\u003eAWS Bedrock Claude Opus\u003c\/b\u003e for final approve\/reject and corrected output when needed.\u003c\/li\u003e\n  \u003cli\u003eLogic and control using \u003cb\u003eif\u003c\/b\u003e, \u003cb\u003eset\u003c\/b\u003e, and \u003cb\u003ecode\u003c\/b\u003e nodes to enforce \u003cb\u003emaxRounds\u003c\/b\u003e\/\u003cb\u003emaxCycles\u003c\/b\u003e.\u003c\/li\u003e\n\u003c\/ul\u003e","brand":"N8N Commerce","offers":[{"title":"Default Title","offer_id":46075752874163,"sku":"N8N-19828","price":10.99,"currency_code":"GBP","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0749\/6279\/6723\/files\/g-YN4yRpXNWRSwMTS7scY_63PzxW5F.png?v=1790154252","url":"https:\/\/buyflowscripts.com\/products\/n8n-workflow-multi-llm-debate-validate-chat-answers","provider":"N8N Commerce","version":"1.0","type":"link"}