Track and Monitor AI Token Usage Metrics for Openai and Gemini Models

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 →

{
  "meta": {
    "templateCredsSetupCompleted": true
  },
  "nodes": [
    {
      "id": "39c42f46-1415-479e-afce-63877743d4c4",
      "name": "Get an execution",
      "type": "n8n-nodes-base.n8n",
      "position": [
        160,
        224
      ],
      "parameters": {
        "options": {
          "activeWorkflows": true
        },
        "resource": "execution",
        "operation": "get",
        "executionId": "={{ $('Start').item.json.excecution_id }}",
        "requestOptions": {}
      },
      "credentials": {
        "n8nApi": {
          "name": "<your credential>"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "01eb07a4-4d4d-4abf-b89b-01eef5b995e1",
      "name": "Start",
      "type": "n8n-nodes-base.executeWorkflowTrigger",
      "position": [
        -64,
        224
      ],
      "parameters": {
        "workflowInputs": {
          "values": [
            {
              "name": "excecution_id"
            },
            {
              "name": "model_names",
              "type": "array"
            }
          ]
        }
      },
      "typeVersion": 1.1
    },
    {
      "id": "2d313b6b-7bf9-4b90-95ac-868642f1bd35",
      "name": "map json",
      "type": "n8n-nodes-base.code",
      "position": [
        384,
        224
      ],
      "parameters": {
        "jsCode": "const input = $input.all();\nconst data = input[0].json;\n\nconst keysToSearch = $('Start').item.json.model_names;\nconst nodes = data.workflowData?.nodes ?? [];\n\nconst totals = {\n\tcompletionTokens: 0,\n\tpromptTokens: 0,\n\ttotalTokens: 0,\n};\n\nconst models = new Set();\nconst detailedUsages = [];\n\nfunction accumulate(usage, model) {\n\tif (\n\t\ttypeof usage.completionTokens === 'number' &&\n\t\ttypeof usage.promptTokens === 'number' &&\n\t\ttypeof usage.totalTokens === 'number'\n\t) {\n\t\ttotals.completionTokens += usage.completionTokens;\n\t\ttotals.promptTokens += usage.promptTokens;\n\t\ttotals.totalTokens += usage.totalTokens;\n\n\t\tif (model) models.add(model);\n\n\t\tdetailedUsages.push({\n\t\t\tmodel,\n\t\t\tusage: { ...usage },\n\t\t});\n\t}\n}\n\nfunction findTokenUsageRecursively(obj, model) {\n\tif (typeof obj !== 'object' || obj === null) return;\n\n\tif (\n\t\ttypeof obj.completionTokens === 'number' &&\n\t\ttypeof obj.promptTokens === 'number' &&\n\t\ttypeof obj.totalTokens === 'number'\n\t) {\n\t\taccumulate(obj, model);\n\t\treturn;\n\t}\n\n\tif (Array.isArray(obj)) {\n\t\tobj.forEach((item) => findTokenUsageRecursively(item, model));\n\t} else {\n\t\tconst detectedModel = obj.model || model;\n\t\tfor (const key in obj) {\n\t\t\tif (Object.prototype.hasOwnProperty.call(obj, key)) {\n\t\t\t\tfindTokenUsageRecursively(obj[key], detectedModel);\n\t\t\t}\n\t\t}\n\t}\n}\n\nfunction processExecution(exec, nodeName) {\n\tconst node = nodes.find((n) => n.name === nodeName);\n\tconst model = node?.parameters?.modelName || node.parameters?.model || node.parameters?.name;\n\tfindTokenUsageRecursively(exec, model);\n}\n\n// Procesar cada nodo indicado\nfor (const key of keysToSearch) {\n\tconst executions = data.data?.resultData?.runData?.[key];\n\tif (!executions) continue;\n\n\tif (Array.isArray(executions)) {\n\t\texecutions.forEach((exec) => processExecution(exec, key));\n\t} else if (typeof executions === 'object') {\n\t\tprocessExecution(executions, key);\n\t}\n}\n\nreturn [\n\t{\n\t\tjson: {\n\t\t\ttotals,\n\t\t\tmodels: Array.from(models),\n\t\t\tdetailedUsages,\n\t\t},\n\t\tpairedItem: [{ item: 0 }] \n\t},\n];\n"
      },
      "typeVersion": 2
    },
    {
      "id": "65843cdf-a21e-428d-b67f-14a5eaba4573",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        880,
        -368
      ],
      "parameters": {
        "width": 448,
        "height": 384,
        "content": "### Sticky Note (Espa\u00f1ol)\n- \u26a0\ufe0f Ejecutar siempre al final del workflow  \n- \u2699\ufe0f Desactivar \u201cEsperar a que termine\u201d  \n- \ud83c\udd94 Requiere `execution_id` y `model_names` (array nodos AI)  \n- \ud83d\udcca Extrae m\u00e9tricas de uso de tokens y modelos usados\n\n---\n\n### Sticky Note (English)\n- \u26a0\ufe0f Always run at the end of the workflow  \n- \u2699\ufe0f Disable \u201cWait until finished\u201d  \n- \ud83c\udd94 Requires `execution_id` and `model_names` (AI node array)  \n- \ud83d\udcca Extracts token usage metrics and used models\n"
      },
      "typeVersion": 1
    },
    {
      "id": "a17a8cdd-2e14-4c41-93cd-e484992ae53b",
      "name": "Execute Workflow",
      "type": "n8n-nodes-base.executeWorkflow",
      "position": [
        624,
        -112
      ],
      "parameters": {
        "options": {
          "waitForSubWorkflow": false
        },
        "workflowId": {
          "__rl": true,
          "mode": "list",
          "value": "6NVerEoYAOMrSF7k",
          "cachedResultName": "workflow shared"
        },
        "workflowInputs": {
          "value": {
            "model_names": "={{ [\"gemini\", \"openai\"] }}",
            "excecution_id": "={{ $execution.id }}"
          },
          "schema": [
            {
              "id": "excecution_id",
              "type": "string",
              "display": true,
              "required": false,
              "displayName": "excecution_id",
              "defaultMatch": false,
              "canBeUsedToMatch": true
            },
            {
              "id": "model_names",
              "type": "array",
              "display": true,
              "required": false,
              "displayName": "model_names",
              "defaultMatch": false,
              "canBeUsedToMatch": true
            }
          ],
          "mappingMode": "defineBelow",
          "matchingColumns": [],
          "attemptToConvertTypes": false,
          "convertFieldsToString": true
        }
      },
      "typeVersion": 1.2
    },
    {
      "id": "28b2d6d1-ea89-48fd-a0b5-930ff7cb130c",
      "name": "When chat message received",
      "type": "@n8n/n8n-nodes-langchain.chatTrigger",
      "position": [
        -64,
        -104
      ],
      "parameters": {
        "options": {}
      },
      "typeVersion": 1.3
    },
    {
      "id": "a9bfcd5e-38f2-414a-80ba-cece8fd2d318",
      "name": "AI Agent",
      "type": "@n8n/n8n-nodes-langchain.agent",
      "position": [
        216,
        -208
      ],
      "parameters": {
        "options": {},
        "needsFallback": true
      },
      "typeVersion": 2.2
    },
    {
      "id": "d6e22621-3dfb-4990-80b3-74651fbeb35e",
      "name": "gemini",
      "type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
      "position": [
        160,
        16
      ],
      "parameters": {
        "options": {}
      },
      "credentials": {
        "googlePalmApi": {
          "name": "<your credential>"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "6243ac56-332d-4444-9597-9e53364e9a76",
      "name": "openai",
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
      "position": [
        288,
        16
      ],
      "parameters": {
        "model": {
          "__rl": true,
          "mode": "list",
          "value": "gpt-4.1-mini"
        },
        "options": {}
      },
      "credentials": {
        "openAiApi": {
          "name": "<your credential>"
        }
      },
      "typeVersion": 1.2
    },
    {
      "id": "61f5c198-03e5-4941-897f-1756432583b7",
      "name": "Simple Memory",
      "type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
      "position": [
        416,
        16
      ],
      "parameters": {},
      "typeVersion": 1.3
    }
  ],
  "connections": {
    "Start": {
      "main": [
        [
          {
            "node": "Get an execution",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "gemini": {
      "ai_languageModel": [
        [
          {
            "node": "AI Agent",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "openai": {
      "ai_languageModel": [
        [
          {
            "node": "AI Agent",
            "type": "ai_languageModel",
            "index": 1
          }
        ]
      ]
    },
    "AI Agent": {
      "main": [
        [
          {
            "node": "Execute Workflow",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "map json": {
      "main": [
        []
      ]
    },
    "Simple Memory": {
      "ai_memory": [
        [
          {
            "node": "AI Agent",
            "type": "ai_memory",
            "index": 0
          }
        ]
      ]
    },
    "Get an execution": {
      "main": [
        [
          {
            "node": "map json",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "When chat message received": {
      "main": [
        [
          {
            "node": "AI Agent",
            "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.

• googlePalmApi
• n8nApi
• openAiApi

About this workflow

📊 Token Usage Metrics Workflow

Source: https://n8n.io/workflows/7265/ — original creator credit. Request a take-down →

More AI & RAG workflows → · Browse all categories →

Related workflows

Workflows that share integrations, category, or trigger type with this one. All free to copy and import.