Yehor EGMS 
This workflow corresponds to n8n.io template #12471 — we link there as the canonical source.
This workflow follows the Chainllm → Chat Trigger recipe pattern — see all workflows that pair these two integrations.
The workflow JSON
Copy or download the full n8n JSON below. Paste it into a new n8n workflow, add your credentials, activate. Full import guide →
{
"id": "FeqLfdT3r8NYWKlC",
"meta": {
"templateCredsSetupCompleted": true
},
"name": "LLM Council",
"tags": [
{
"id": "k0gyhK57BQktK4yA",
"name": "Technical",
"createdAt": "2025-09-04T10:44:14.340Z",
"updatedAt": "2025-09-04T11:13:29.041Z"
}
],
"nodes": [
{
"id": "8cf3f133-6f14-43e6-91f1-5949d44b230f",
"name": "When chat message received",
"type": "@n8n/n8n-nodes-langchain.chatTrigger",
"position": [
-1552,
64
],
"parameters": {
"options": {}
},
"typeVersion": 1.4
},
{
"id": "b142fc0d-12d3-4f3a-8114-fe56c09c894f",
"name": "Basic LLM Chain",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-1216,
-448
],
"parameters": {
"batching": {},
"messages": {
"messageValues": [
{
"message": "Give an answer in English"
}
]
}
},
"typeVersion": 1.7
},
{
"id": "b95480e3-4fcd-4bfe-873b-dd6718fc2aeb",
"name": "Basic LLM Chain1",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-1216,
-96
],
"parameters": {
"batching": {},
"messages": {
"messageValues": [
{
"message": "Give an answer in English"
}
]
}
},
"typeVersion": 1.7
},
{
"id": "f600cb1b-c72f-4fe1-a477-fe0c84a7dc63",
"name": "Basic LLM Chain2",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-1216,
240
],
"parameters": {
"batching": {},
"messages": {
"messageValues": [
{
"message": "Give an answer in English"
}
]
}
},
"typeVersion": 1.7
},
{
"id": "c2a852dc-9201-4224-baf5-4a1862b5b6fb",
"name": "Basic LLM Chain3",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-1216,
576
],
"parameters": {
"batching": {},
"messages": {
"messageValues": [
{
"message": "Give an answer in English"
}
]
}
},
"typeVersion": 1.7
},
{
"id": "070699ee-c2db-4676-bd74-3447a9f66f5c",
"name": "Merge",
"type": "n8n-nodes-base.merge",
"position": [
-208,
512
],
"parameters": {
"mode": "combine",
"options": {},
"combineBy": "combineByPosition",
"numberInputs": 4
},
"typeVersion": 3.2
},
{
"id": "03c5d08a-eb6c-46d8-97a7-abd34ec3f92e",
"name": "Merge1",
"type": "n8n-nodes-base.merge",
"position": [
-208,
192
],
"parameters": {
"mode": "combine",
"options": {},
"combineBy": "combineByPosition",
"numberInputs": 4
},
"typeVersion": 3.2
},
{
"id": "e224a52d-0041-4abe-94f7-3f827832ca67",
"name": "Merge3",
"type": "n8n-nodes-base.merge",
"position": [
-208,
-432
],
"parameters": {
"mode": "combine",
"options": {},
"combineBy": "combineByPosition",
"numberInputs": 4
},
"typeVersion": 3.2
},
{
"id": "dea776a4-ecc1-4a5d-a72a-3b7cde77893d",
"name": "gemini1",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-1216,
736
],
"parameters": {
"model": "google/gemini-3-flash-preview",
"options": {
"maxTokens": 4000,
"temperature": 0.7
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "635fe200-5ca0-44a5-931a-f688bcbdc662",
"name": "Merge4",
"type": "n8n-nodes-base.merge",
"position": [
608,
32
],
"parameters": {
"mode": "combine",
"options": {},
"combineBy": "combineByPosition",
"numberInputs": 4
},
"typeVersion": 3.2
},
{
"id": "674fd520-4610-439a-982e-99d1b1d58bb6",
"name": "Basic LLM Chain8",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
1088,
64
],
"parameters": {
"text": "=You are the Chairman of an LLM Council. Multiple AI models have provided responses to a user's question, and then ranked each other's responses.\n\nOriginal Question: {{ $('When chat message received').item.json.chatInput }}\n\nSTAGE 1 - Individual Responses:\nModel: anthropic/claud\nResponse: {{ $('response a').item.json.text_A }}\n\nModel: openai/gpt\nResponse: {{ $('response b').item.json.text_B }}\n\nModel: x-ai/grok\nResponse: {{ $('response c').item.json.text_C }}\n\nModel: google/gemini\nResponse: {{ $('response d').item.json.text_D }}\n\nSTAGE 2 - Peer Rankings:\nModel: anthropic/claud\nRanking: {{ $('evaluate_a').item.json.Raiting_1 }}\n\nModel: openai/gpt\nRanking: {{ $('evaluate_b').item.json.Raiting_2 }}\n\nModel: x-ai/grok\nRanking: {{ $('evaluate_c').item.json.Raiting_3 }}\n\nModel: google/gemini\nRanking: {{ $('evaluate_d').item.json.Raiting_4 }}\n\nYour task as Chairman is to synthesize all of this information into a single, comprehensive, accurate answer to the user's original question. \n\nConsider:\n- The individual responses and their insights\n- The peer rankings and what they reveal about response quality\n- Any patterns of agreement or disagreement\n\nProvide a clear, well-reasoned final answer that represents the council's collective wisdom:",
"batching": {},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "c1355eee-64bc-4277-8e76-5b38ce176b9e",
"name": "response a",
"type": "n8n-nodes-base.set",
"position": [
-896,
-448
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "d1ddeb46-5ee6-4bd5-9a38-d7fd7b1b193e",
"name": "text_A",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "69ea4488-e720-4933-9e12-6870c5ba0b06",
"name": "response b",
"type": "n8n-nodes-base.set",
"position": [
-896,
-96
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "c7b84760-65b0-40bf-b516-055096148e80",
"name": "text_B",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "7db542b5-781c-4f39-936d-ed1dfad6c042",
"name": "response c",
"type": "n8n-nodes-base.set",
"position": [
-896,
240
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "7e5f4c68-c155-4b7f-988a-678153b7dc88",
"name": "text_C",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "72e0bf78-e133-41d7-8825-8edf9c9d7d6b",
"name": "response d",
"type": "n8n-nodes-base.set",
"position": [
-896,
576
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "c2f263f1-94d0-4e9f-857d-e103858d7a3b",
"name": "text_D",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "5b505dd2-decf-4f3e-be31-83017fcb0e78",
"name": "Code in JavaScript",
"type": "n8n-nodes-base.code",
"position": [
816,
64
],
"parameters": {
"jsCode": "const itemData = items[0].json; // \u2190 \u041f\u0420\u0410\u0412\u0418\u041b\u042c\u041d\u041e!\n\nfunction parseRanking(text) {\n const ranking = {};\n const lines = text.split('\\n');\n let inFinalRanking = false;\n \n lines.forEach(line => {\n if (line.includes('FINAL RANKING')) {\n inFinalRanking = true;\n }\n if (inFinalRanking && line.trim().length > 0) {\n const match = line.match(/(\\d+)\\.\\s*(Response [A-D])/);\n if (match) {\n ranking[match[2]] = parseInt(match[1]);\n }\n }\n });\n \n return ranking;\n}\n\nconst allRankings = [];\n['Raiting_1', 'Raiting_2', 'Raiting_3', 'Raiting_4'].forEach(key => {\n if (itemData[key]) {\n const parsed = parseRanking(itemData[key]);\n allRankings.push(parsed);\n console.log(`Parsed ${key}:`, parsed);\n }\n});\n\nconst aggregated = {};\n['Response A', 'Response B', 'Response C', 'Response D'].forEach(resp => {\n aggregated[resp] = { positions: [] };\n});\n\nallRankings.forEach(ranking => {\n Object.entries(ranking).forEach(([response, position]) => {\n if (aggregated[response]) {\n aggregated[response].positions.push(position);\n }\n });\n});\n\nObject.keys(aggregated).forEach(response => {\n const positions = aggregated[response].positions;\n if (positions.length > 0) {\n const avg = positions.reduce((a,b) => a+b, 0) / positions.length;\n aggregated[response].average = parseFloat(avg.toFixed(2));\n aggregated[response].count = positions.length;\n aggregated[response].total = positions.reduce((a,b) => a+b, 0);\n }\n});\n\nconst sorted = Object.entries(aggregated)\n .sort((a, b) => a[1].average - b[1].average)\n .reduce((obj, [key, val]) => { obj[key] = val; return obj; }, {});\n\nreturn [{\n json: {\n raw_rankings: allRankings,\n aggregated_rankings: sorted,\n best_response: Object.keys(sorted)[0],\n best_score: sorted[Object.keys(sorted)[0]].average,\n rankings_count: allRankings.length\n }\n}];\n"
},
"typeVersion": 2
},
{
"id": "9bb3a92e-55cc-480d-b32f-7c2d098071bd",
"name": "claude3",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-1216,
-288
],
"parameters": {
"model": "anthropic/claude-sonnet-4.5",
"options": {
"maxTokens": 4000,
"temperature": 0.7
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "100c79d2-8d3f-40f2-8a79-5cbc347eac81",
"name": "openAI2",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-1216,
80
],
"parameters": {
"model": "openai/gpt-5.1",
"options": {
"maxTokens": 4000,
"temperature": 0.7
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "f695dfab-5eb3-4d70-8112-7a3ccd2cb6b5",
"name": "groq1",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-1216,
400
],
"parameters": {
"model": "x-ai/grok-4",
"options": {
"maxTokens": 4000,
"temperature": 0.7
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "13769759-f6e6-4030-bd40-f447239d4d71",
"name": "Merge5",
"type": "n8n-nodes-base.merge",
"position": [
-208,
-112
],
"parameters": {
"mode": "combine",
"options": {},
"combineBy": "combineByPosition",
"numberInputs": 4
},
"typeVersion": 3.2
},
{
"id": "894a1d5e-b727-4ddb-8631-6988d9f403df",
"name": "claude4",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
1088,
224
],
"parameters": {
"model": "google/gemini-3-pro-preview",
"options": {
"temperature": 0.6
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "e2bd5116-4489-46ee-9108-b641446301d8",
"name": "evaluate claude",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-16,
-400
],
"parameters": {
"text": "=You are evaluating different responses to the following question:\n\nQuestion: {{ $('When chat message received').item.json.chatInput }}\n\nHere are the responses from different models (anonymized):\n\nResponse A:\n{{ $json.text_A }}\n\nResponse B:\n{{ $json.text_B }}\n\nResponse C:\n{{ $json.text_C }}\n\nResponse D:\n{{ $json.text_D }}\n\nYour task:\n1. First, evaluate each response individually. For each response, explain what it does well and what it does poorly.\n2. Then, at the very end of your response, provide a final ranking.\n\nIMPORTANT: Your final ranking MUST be formatted EXACTLY as follows:\n- Start with the line \"FINAL RANKING:\" (all caps, with colon)\n- Then list the responses from best to worst as a numbered list\n- Each line should be: number, period, space, then ONLY the response label (e.g., \"1. Response A\")\n- Do not add any other text or explanations in the ranking section\n- If there is an error with the data, please report it immediately and do not generate any response.\n- give answer on english\n\n\nExample of the correct format for your ENTIRE response:\n\nResponse A provides good detail on X but misses Y...\nResponse B is accurate but lacks depth on Z...\nResponse C offers the most comprehensive answer...\n\nFINAL RANKING:\n1. Response C\n2. Response A\n3. Response B\n4. Response D\n\nNow provide your evaluation and ranking:\n\n",
"batching": {},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "de6f093f-e113-4028-8ff5-d99ed5cbf97a",
"name": "evaluate grok",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-16,
224
],
"parameters": {
"text": "=You are evaluating different responses to the following question:\n\nQuestion: {{ $('When chat message received').item.json.chatInput }}\n\nHere are the responses from different models (anonymized):\n\nResponse A:\n{{ $json.text_A }}\n\nResponse B:\n{{ $json.text_B }}\n\nResponse C:\n{{ $json.text_C }}\n\nResponse D:\n{{ $json.text_D }}\n\nYour task:\n1. First, evaluate each response individually. For each response, explain what it does well and what it does poorly.\n2. Then, at the very end of your response, provide a final ranking.\n\nIMPORTANT: Your final ranking MUST be formatted EXACTLY as follows:\n- Start with the line \"FINAL RANKING:\" (all caps, with colon)\n- Then list the responses from best to worst as a numbered list\n- Each line should be: number, period, space, then ONLY the response label (e.g., \"1. Response A\")\n- Do not add any other text or explanations in the ranking section\n- If there is an error with the data, please report it immediately and do not generate any response.\n- give answer on english\n\n\nExample of the correct format for your ENTIRE response:\n\nResponse A provides good detail on X but misses Y...\nResponse B is accurate but lacks depth on Z...\nResponse C offers the most comprehensive answer...\n\nFINAL RANKING:\n1. Response C\n2. Response A\n3. Response B\n4. Response D\n\nNow provide your evaluation and ranking:\n\n",
"batching": {},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "26468bc2-7162-497c-8563-847e8e84d5db",
"name": "evaluate gemini",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-16,
544
],
"parameters": {
"text": "=You are evaluating different responses to the following question:\n\nQuestion: {{ $('When chat message received').item.json.chatInput }}\n\nHere are the responses from different models (anonymized):\n\nResponse A:\n{{ $json.text_A }}\n\nResponse B:\n{{ $json.text_B }}\n\nResponse C:\n{{ $json.text_C }}\n\nResponse D:\n{{ $json.text_D }}\n\nYour task:\n1. First, evaluate each response individually. For each response, explain what it does well and what it does poorly.\n2. Then, at the very end of your response, provide a final ranking.\n\nIMPORTANT: Your final ranking MUST be formatted EXACTLY as follows:\n- Start with the line \"FINAL RANKING:\" (all caps, with colon)\n- Then list the responses from best to worst as a numbered list\n- Each line should be: number, period, space, then ONLY the response label (e.g., \"1. Response A\")\n- Do not add any other text or explanations in the ranking section\n- If there is an error with the data, please report it immediately and do not generate any response.\n- give answer on english\n\n\nExample of the correct format for your ENTIRE response:\n\nResponse A provides good detail on X but misses Y...\nResponse B is accurate but lacks depth on Z...\nResponse C offers the most comprehensive answer...\n\nFINAL RANKING:\n1. Response C\n2. Response A\n3. Response B\n4. Response D\n\nNow provide your evaluation and ranking:\n\n",
"batching": {},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "365bc7bc-b8d8-47c1-8169-6eddecbebbdd",
"name": "evaluate gpt",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-16,
-80
],
"parameters": {
"text": "=You are evaluating different responses to the following question:\n\nQuestion: {{ $('When chat message received').item.json.chatInput }}\n\nHere are the responses from different models (anonymized):\n\nResponse A:\n{{ $json.text_A }}\n\nResponse B:\n{{ $json.text_B }}\n\nResponse C:\n{{ $json.text_C }}\n\nResponse D:\n{{ $json.text_D }}\n\nYour task:\n1. First, evaluate each response individually. For each response, explain what it does well and what it does poorly.\n2. Then, at the very end of your response, provide a final ranking.\n\nIMPORTANT: Your final ranking MUST be formatted EXACTLY as follows:\n- Start with the line \"FINAL RANKING:\" (all caps, with colon)\n- Then list the responses from best to worst as a numbered list\n- Each line should be: number, period, space, then ONLY the response label (e.g., \"1. Response A\")\n- Do not add any other text or explanations in the ranking section\n- If there is an error with the data, please report it immediately and do not generate any response.\n- give answer on english\n\n\nExample of the correct format for your ENTIRE response:\n\nResponse A provides good detail on X but misses Y...\nResponse B is accurate but lacks depth on Z...\nResponse C offers the most comprehensive answer...\n\nFINAL RANKING:\n1. Response C\n2. Response A\n3. Response B\n4. Response D\n\nNow provide your evaluation and ranking:\n\n",
"batching": {},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "ae986c1c-e2c5-456e-98c7-b6d5018acb2e",
"name": "evaluate_a",
"type": "n8n-nodes-base.set",
"position": [
304,
-400
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "6c544736-80c2-4f52-b4b9-5236f28f524c",
"name": "Raiting_1",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "c03b9556-9260-4776-8498-4d971a447c28",
"name": "evaluate_b",
"type": "n8n-nodes-base.set",
"position": [
304,
-80
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "5df1b7b3-57e0-43a0-9214-70739d8cea8a",
"name": "Raiting_2",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "449c03be-775b-42cd-b329-d021666a10c6",
"name": "evaluate_c",
"type": "n8n-nodes-base.set",
"position": [
304,
224
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "21e8321a-1317-4e5b-bfdb-38947be54ee2",
"name": "Raiting_3",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "81291e94-9d74-4be7-b801-d52dadb97a33",
"name": "evaluate_d",
"type": "n8n-nodes-base.set",
"position": [
304,
544
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "d1f431e9-7a2a-4956-b134-ec439d956109",
"name": "Raiting_4",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "e10bd59b-df65-4f06-b01c-12e5c9718ee5",
"name": "claude",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-16,
-240
],
"parameters": {
"model": "anthropic/claude-sonnet-4.5",
"options": {
"maxTokens": 2000,
"temperature": 0.3
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "e327a8cb-6eb9-4387-869f-1ebfd2bd8035",
"name": "openAI",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-16,
96
],
"parameters": {
"model": "openai/gpt-5.1",
"options": {
"maxTokens": 2000,
"temperature": 0.3
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "716101d0-c2a8-4266-8931-1d7805cd1152",
"name": "groq",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-16,
400
],
"parameters": {
"model": "x-ai/grok-4",
"options": {
"maxTokens": 2000,
"temperature": 0.3
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "375d1fe2-f040-498f-806d-dbf40a123e99",
"name": "gemini2",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
"position": [
-16,
704
],
"parameters": {
"model": "google/gemini-3-flash-preview",
"options": {
"maxTokens": 2000,
"temperature": 0.3
}
},
"credentials": {
"openRouterApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "80fa8e7d-4e00-48f7-b608-7cc55e55b5b0",
"name": "Sticky Note2",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1232,
-656
],
"parameters": {
"color": 6,
"width": 256,
"height": 1552,
"content": "## Stage 1\nGet answers from different models"
},
"typeVersion": 1
},
{
"id": "7b289f3b-689e-40b7-b227-8dd61a59a27d",
"name": "Sticky Note3",
"type": "n8n-nodes-base.stickyNote",
"position": [
-960,
-656
],
"parameters": {
"color": 6,
"height": 1552,
"content": "## Stage 2\nFor convenience and impartiality, we store and anonymize responses."
},
"typeVersion": 1
},
{
"id": "76699f65-6b7c-43d7-a4c6-bf73dc3f18d2",
"name": "Sticky Note4",
"type": "n8n-nodes-base.stickyNote",
"position": [
-288,
-656
],
"parameters": {
"color": 6,
"height": 1552,
"content": "## For convenience and impartiality, we store and anonymize responses."
},
"typeVersion": 1
},
{
"id": "3e03a8aa-ce45-4db2-906d-a83c28e6b05e",
"name": "Sticky Note5",
"type": "n8n-nodes-base.stickyNote",
"position": [
-32,
-656
],
"parameters": {
"color": 6,
"width": 256,
"height": 1552,
"content": "## Stage 3\nFor convenience and impartiality, we store and anonymize responses."
},
"typeVersion": 1
},
{
"id": "509b338b-6145-465f-ae89-691dd4268cac",
"name": "Sticky Note6",
"type": "n8n-nodes-base.stickyNote",
"position": [
240,
-656
],
"parameters": {
"color": 6,
"height": 1552,
"content": "## we get an analysis of all the answers"
},
"typeVersion": 1
},
{
"id": "85243108-8cd7-4ea6-8052-4bee850b19b4",
"name": "Sticky Note7",
"type": "n8n-nodes-base.stickyNote",
"position": [
752,
-96
],
"parameters": {
"color": 6,
"height": 432,
"content": "## Stage 4\nThis is where the magic happens and all the scores are ranked."
},
"typeVersion": 1
},
{
"id": "32e6d9fe-1f5b-44f5-8e9b-392b5dcd24e3",
"name": "Sticky Note8",
"type": "n8n-nodes-base.stickyNote",
"position": [
1040,
-96
],
"parameters": {
"color": 6,
"width": 304,
"height": 432,
"content": "## Stage 5\nAll responses are evaluated by one model and a generalized and summarized response is provided."
},
"typeVersion": 1
},
{
"id": "26f6efd1-99be-4851-bc4c-ee907062e511",
"name": "Anthropic Chat Model",
"type": "@n8n/n8n-nodes-langchain.lmChatAnthropic",
"position": [
-1680,
704
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "claude-sonnet-4-5-20250929",
"cachedResultName": "Claude Sonnet 4.5"
},
"options": {}
},
"credentials": {
"anthropicApi": {
"name": "<your credential>"
}
},
"typeVersion": 1.3
},
{
"id": "e7ff77c9-6ae6-45af-b761-973dffb7612f",
"name": "xAI Grok Chat Model",
"type": "@n8n/n8n-nodes-langchain.lmChatXAiGrok",
"position": [
-1520,
704
],
"parameters": {
"options": {}
},
"credentials": {
"xAiApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "d4e47a9a-122e-4e0d-933c-142054d0a0eb",
"name": "OpenAI Chat Model",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
-1520,
544
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4.1-mini"
},
"options": {},
"builtInTools": {}
},
"credentials": {
"openAiApi": {
"name": "<your credential>"
}
},
"typeVersion": 1.3
},
{
"id": "7f89855e-ba74-4fd7-b0a5-131baf63f03e",
"name": "Google Gemini Chat Model",
"type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
"position": [
-1680,
544
],
"parameters": {
"options": {}
},
"credentials": {
"googlePalmApi": {
"name": "<your credential>"
}
},
"typeVersion": 1
},
{
"id": "2a7ee75f-d680-425e-9916-f66ec75be81c",
"name": "Sticky Note9",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1712,
432
],
"parameters": {
"color": 7,
"width": 352,
"height": 416,
"content": "## possible options for data analysis directly"
},
"typeVersion": 1
},
{
"id": "e36c9d6b-dd88-4b09-9a46-63fe2783db9f",
"name": "Sticky Note",
"type": "n8n-nodes-base.stickyNote",
"position": [
-2144,
432
],
"parameters": {
"color": 7,
"width": 416,
"height": 416,
"content": "## possible data entry options"
},
"typeVersion": 1
},
{
"id": "1848ca5f-2c79-4d4e-865f-90e4363e9ba0",
"name": "Send a message1",
"type": "n8n-nodes-base.gmail",
"position": [
-1920,
512
],
"parameters": {
"options": {}
},
"credentials": {
"gmailOAuth2": {
"name": "<your credential>"
}
},
"typeVersion": 2.2
},
{
"id": "2e723348-7986-4097-a164-f41ca166cd4d",
"name": "Send a message",
"type": "n8n-nodes-base.slack",
"position": [
-1920,
688
],
"parameters": {
"otherOptions": {}
},
"typeVersion": 2.4
},
{
"id": "0c047a14-3ccf-456a-83ac-0f616cbcc5e4",
"name": "Send message",
"type": "n8n-nodes-base.whatsApp",
"position": [
-2112,
688
],
"parameters": {
"operation": "send",
"additionalFields": {}
},
"typeVersion": 1.1
},
{
"id": "e75a552e-82b5-4eba-b201-93586e88ff79",
"name": "Send a text message",
"type": "n8n-nodes-base.telegram",
"position": [
-2112,
512
],
"parameters": {
"additionalFields": {}
},
"credentials": {
"telegramApi": {
"name": "<your credential>"
}
},
"typeVersion": 1.2
},
{
"id": "c86237dd-e0a6-4228-9dd0-1ab04bf1fdcd",
"name": "Sticky Note10",
"type": "n8n-nodes-base.stickyNote",
"position": [
-2960,
-352
],
"parameters": {
"width": 784,
"height": 576,
"content": "## How It Works\n\n1\ufe0f\u20e3 User Input\nA user sends a single question via chat. This message becomes the shared input for the entire workflow.\n\n2\ufe0f\u20e3 Parallel LLM Responses\nThe same question is sent simultaneously to multiple large language models.\nEach model generates its own answer independently, using the same prompt but its own reasoning.\n\n3\ufe0f\u20e3 Response Anonymization\nAll model outputs are stored as anonymous responses (Response A\u2013D).\nThis prevents evaluator models from knowing which LLM produced which answer.\n\n4\ufe0f\u20e3 Peer Evaluation & Ranking\nEach LLM reviews all anonymized responses, analyzes their strengths and weaknesses, and produces a strict ranking from best to worst.\n\n5\ufe0f\u20e3 Ranking Aggregation\nAll rankings are collected and aggregated.\nThe workflow calculates average positions and identifies the strongest response based on collective evaluation.\n\n6\ufe0f\u20e3 Final Consensus Answer\nA dedicated \u201cChairman\u201d model reviews all responses and rankings, detects agreement and disagreement patterns, and synthesizes one clear, high-quality final answer.\n\n\u27a1\ufe0f Result: a consensus-driven response built from multiple independent LLM perspectives, not a single-model output.\n\n"
},
"typeVersion": 1
},
{
"id": "d129dc4f-0ec2-46b2-8b83-e4bec8426df2",
"name": "Sticky Note1",
"type": "n8n-nodes-base.stickyNote",
"position": [
-2960,
256
],
"parameters": {
"color": 3,
"width": 784,
"height": 592,
"content": "## Setup Steps\n1\ufe0f\u20e3 Import the Workflow\nImport the provided n8n workflow template into your instance.\n\n2\ufe0f\u20e3 Configure OpenRouter Credentials\nAdd your OpenRouter API key in n8n credentials and connect it to all LLM nodes in the workflow.\n\n3\ufe0f\u20e3 Select LLM Models\nChoose which models will participate in the council (e.g. GPT, Claude, Gemini, Grok).\nYou can add or remove models without changing the core logic.\n\n4\ufe0f\u20e3 Review Prompt Nodes\nCheck the system and evaluation prompts.\nAdjust tone, language, or strictness if your use case requires it.\n\n5\ufe0f\u20e3 Verify the Aggregation Code Node\nEnsure the JavaScript aggregation node matches the number of active LLMs and expected ranking format.\n\n6\ufe0f\u20e3 Test with a Sample Question\nRun the workflow with a simple input to confirm that:\n\nAll models respond\n\nRankings are generated\n\nA final consensus answer is produced\n\n7\ufe0f\u20e3 Activate the Workflow\nOnce verified, activate the workflow and use it in production or connect it to other automations."
},
"typeVersion": 1
}
],
"active": false,
"settings": {
"executionOrder": "v1"
},
"versionId": "d50be5bd-4366-46ca-a785-c924dde43b65",
"connections": {
"groq": {
"ai_languageModel": [
[
{
"node": "evaluate grok",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Merge": {
"main": [
[
{
"node": "evaluate gemini",
"type": "main",
"index": 0
}
]
]
},
"groq1": {
"ai_languageModel": [
[
{
"node": "Basic LLM Chain2",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Merge1": {
"main": [
[
{
"node": "evaluate grok",
"type": "main",
"index": 0
}
]
]
},
"Merge3": {
"main": [
[
{
"node": "evaluate claude",
"type": "main",
"index": 0
}
]
]
},
"Merge4": {
"main": [
[
{
"node": "Code in JavaScript",
"type": "main",
"index": 0
}
]
]
},
"Merge5": {
"main": [
[
{
"node": "evaluate gpt",
"type": "main",
"index": 0
}
]
]
},
"claude": {
"ai_languageModel": [
[
{
"node": "evaluate claude",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"openAI": {
"ai_languageModel": [
[
{
"node": "evaluate gpt",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"claude3": {
"ai_languageModel": [
[
{
"node": "Basic LLM Chain",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"claude4": {
"ai_languageModel": [
[
{
"node": "Basic LLM Chain8",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"gemini1": {
"ai_languageModel": [
[
{
"node": "Basic LLM Chain3",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"gemini2": {
"ai_languageModel": [
[
{
"node": "evaluate gemini",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"openAI2": {
"ai_languageModel": [
[
{
"node": "Basic LLM Chain1",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"evaluate_a": {
"main": [
[
{
"node": "Merge4",
"type": "main",
"index": 0
}
]
]
},
"evaluate_b": {
"main": [
[
{
"node": "Merge4",
"type": "main",
"index": 1
}
]
]
},
"evaluate_c": {
"main": [
[
{
"node": "Merge4",
"type": "main",
"index": 2
}
]
]
},
"evaluate_d": {
"main": [
[
{
"node": "Merge4",
"type": "main",
"index": 3
}
]
]
},
"response a": {
"main": [
[
{
"node": "Merge3",
"type": "main",
"index": 0
},
{
"node": "Merge1",
"type": "main",
"index": 0
},
{
"node": "Merge",
"type": "main",
"index": 0
},
{
"node": "Merge5",
"type": "main",
"index": 0
}
]
]
},
"response b": {
"main": [
[
{
"node": "Merge3",
"type": "main",
"index": 1
},
{
"node": "Merge5",
"type": "main",
"index": 1
},
{
"node": "Merge1",
"type": "main",
"index": 1
},
{
"node": "Merge",
"type": "main",
"index": 1
}
]
]
},
"response c": {
"main": [
[
{
"node": "Merge3",
"type": "main",
"index": 2
},
{
"node": "Merge5",
"type": "main",
"index": 2
},
{
"node": "Merge1",
"type": "main",
"index": 2
},
{
"node": "Merge",
"type": "main",
"index": 2
}
]
]
},
"response d": {
"main": [
[
{
"node": "Merge3",
"type": "main",
"index": 3
},
{
"node": "Merge",
"type": "main",
"index": 3
},
{
"node": "Merge1",
"type": "main",
"index": 3
},
{
"node": "Merge5",
"type": "main",
"index": 3
}
]
]
},
"evaluate gpt": {
"main": [
[
{
"node": "evaluate_b",
"type": "main",
"index": 0
}
]
]
},
"evaluate grok": {
"main": [
[
{
"node": "evaluate_c",
"type": "main",
"index": 0
}
]
]
},
"Basic LLM Chain": {
"main": [
[
{
"node": "response a",
"type": "main",
"index": 0
}
]
]
},
"evaluate claude": {
"main": [
[
{
"node": "evaluate_a",
"type": "main",
"index": 0
}
]
]
},
"evaluate gemini": {
"main": [
[
{
"node": "evaluate_d",
"type": "main",
"index": 0
}
]
]
},
"Basic LLM Chain1": {
"main": [
[
{
"node": "response b",
"type": "main",
"index": 0
}
]
]
},
"Basic LLM Chain2": {
"main": [
[
{
"node": "response c",
"type": "main",
"index": 0
}
]
]
},
"Basic LLM Chain3": {
"main": [
[
{
"node": "response d",
"type": "main",
"index": 0
}
]
]
},
"Code in JavaScript": {
"main": [
[
{
"node": "Basic LLM Chain8",
"type": "main",
"index": 0
}
]
]
},
"When chat message received": {
"main": [
[
{
"node": "Basic LLM Chain",
"type": "main",
"index": 0
},
{
"node": "Basic LLM Chain1",
"type": "main",
"index": 0
},
{
"node": "Basic LLM Chain2",
"type": "main",
"index": 0
},
{
"node": "Basic LLM Chain3",
"type": "main",
"index": 0
}
]
]
}
}
}
Credentials you'll need
Each integration node will prompt for credentials when you import. We strip credential IDs before publishing — you'll add your own.
anthropicApigmailOAuth2googlePalmApiopenAiApiopenRouterApitelegramApixAiApi
Pro
For the full experience including quality scoring and batch install features for each workflow upgrade to Pro
About this workflow
The original LLM Council concept was introduced by Andrej Karpathy and published as an open-source repository demonstrating multi-model consensus and ranking. This workflow is my adaptation of that original idea, reimplemented and structured as a production-ready n8n template.…
Source: https://n8n.io/workflows/12471/ — original creator credit. Request a take-down →
Related workflows
Workflows that share integrations, category, or trigger type with this one. All free to copy and import.
This workflow is designed to intelligently route user queries to the most suitable large language model (LLM) based on the type of request received in a chat environment. It uses structured classifica
Chat Trigger, Agent, Model Selector +7
A Telegram bot that converts natural-language work descriptions into detailed cost estimates using AI parsing, vector search, and the open-source DDC CWICR database with 55,000+ construction work item
HTTP Request, Telegram, Telegram Trigger +6
My workflow 53. Uses formTrigger, httpRequest, lmChatOpenAi, form. Event-driven trigger; 74 nodes.
Form Trigger, HTTP Request, OpenAI Chat +15
Episode 23: UGC with nanobanana. Uses lmChatOpenAi, lmChatOllama, lmChatDeepSeek, lmChatOpenRouter. Event-driven trigger; 74 nodes.
OpenAI Chat, Ollama Chat, Lm Chat Deep Seek +12
This workflow is for beauty salons who want consistent, high‑quality social media content without writing every post manually. It also suits agencies and automation builders who manage multiple beauty
Telegram, Google Sheets Trigger, Agent +26