Generate TikTok Videos with GPT-4o & Telegram

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
  },
  "name": "TikTok Videos That Resonate With Your Target Audience",
  "tags": [
    {
      "name": "Content Creation"
    },
    {
      "name": "AI"
    },
    {
      "name": "TikTok"
    },
    {
      "name": "Marketing"
    }
  ],
  "nodes": [
    {
      "id": "sticky-main-overview",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        0,
        0
      ],
      "parameters": {
        "color": 4,
        "width": 380,
        "height": 904,
        "content": "## TikTok Videos That Resonate With Your Target Audience\nCreate TikTok videos grounded in real human insight using AI and Digital Twins.\n\n### How it works\n1. **Telegram** triggers the workflow (send your product/concept)\n2. **AI Agent** researches trends, gathers audience feedback, writes scripts informed by real insights, and tests them with Digital Twins\n3. **Veo 3** generates video from the most resonant script\n4. **Telegram** sends the video back to you\n\n### Setup\n1. Edit **Workflow Configuration** with:\n   - Product details (name, description, audience, tone)\n2. Connect **Telegram Bot** credentials (@BotFather)\n3. Connect **OpenAI API** credentials on Chat Model node\n4. Connect **Header Auth** on Apify MCP node:\n   - Header: `Authorization`\n   - Value: `Bearer YOUR_APIFY_TOKEN`\n5. Connect **Header Auth** on OriginalVoices node:\n   - Header: `X-Api-Key`\n   - Value: `YOUR_API_KEY`\n6. Connect **Header Auth** on Veo 3 & Poll nodes:\n   - Header: `Authorization`\n   - Value: `Key YOUR_FAL_API_KEY`\n7. **Activate** the workflow\n\n### Customize\n- **Audience**: Refine targetAudience for specific demographics\n- **Trends**: Modify agent prompt for different niches\n- **Scripts**: Adjust number of variations\n- **Video style**: Update Veo 3 prompt\n\n### Need help?\nReach out on Discord (`vedad27`) or email `vedad@originalvoices.ai`"
      },
      "typeVersion": 1
    },
    {
      "id": "sticky-config",
      "name": "Sticky Note Config",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        432,
        112
      ],
      "parameters": {
        "color": 7,
        "width": 228,
        "height": 288,
        "content": "### 1. Trigger"
      },
      "typeVersion": 1
    },
    {
      "id": "sticky-trigger",
      "name": "Sticky Note Trigger",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        704,
        112
      ],
      "parameters": {
        "color": 7,
        "height": 288,
        "content": "### 2. Configuration\nEdit with your product details"
      },
      "typeVersion": 1
    },
    {
      "id": "sticky-agent",
      "name": "Sticky Note Agent",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        960,
        112
      ],
      "parameters": {
        "color": 7,
        "width": 496,
        "height": 484,
        "content": "### 3. Research, Write & Test\nConnect Apify + OriginalVoices credentials"
      },
      "typeVersion": 1
    },
    {
      "id": "sticky-video",
      "name": "Sticky Note Video",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1504,
        112
      ],
      "parameters": {
        "color": 7,
        "width": 420,
        "height": 480,
        "content": "### 4. Generate Video"
      },
      "typeVersion": 1
    },
    {
      "id": "sticky-report",
      "name": "Sticky Note Report",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1984,
        112
      ],
      "parameters": {
        "color": 7,
        "width": 420,
        "height": 480,
        "content": "### 5. Report back to Telegram"
      },
      "typeVersion": 1
    },
    {
      "id": "config-node",
      "name": "Workflow Configuration",
      "type": "n8n-nodes-base.set",
      "position": [
        784,
        208
      ],
      "parameters": {
        "options": {},
        "assignments": {
          "assignments": [
            {
              "id": "product-name",
              "name": "productName",
              "type": "string",
              "value": "Your Product Name"
            },
            {
              "id": "product-description",
              "name": "productDescription",
              "type": "string",
              "value": "A brief description of what your product does and the problem it solves"
            },
            {
              "id": "target-audience",
              "name": "targetAudience",
              "type": "string",
              "value": "US, 18-35, interested in tech and productivity"
            },
            {
              "id": "tone",
              "name": "tone",
              "type": "string",
              "value": "casual, energetic, Gen-Z friendly"
            }
          ]
        }
      },
      "typeVersion": 3.4
    },
    {
      "id": "telegram-trigger",
      "name": "Telegram Trigger",
      "type": "n8n-nodes-base.telegramTrigger",
      "position": [
        496,
        208
      ],
      "parameters": {
        "updates": [
          "message"
        ],
        "additionalFields": {}
      },
      "typeVersion": 1.1
    },
    {
      "id": "creative-agent",
      "name": "Research, Write & Test",
      "type": "@n8n/n8n-nodes-langchain.agent",
      "position": [
        992,
        208
      ],
      "parameters": {
        "text": "=You are a TikTok content strategist who creates viral videos. You have access to two tools:\n\n1. **Apify** - Use this to scrape TikTok for current trending content\n2. **ask_twins** - Use this to get real human feedback (Digital Twins)\n\n---\n\nVIDEO TOPIC/DIRECTION (from user):\n{{ $('Telegram Trigger').item.json.message.text }}\n\nPRODUCT INFO:\n- Name: {{ $('Workflow Configuration').item.json.productName }}\n- Description: {{ $('Workflow Configuration').item.json.productDescription }}\n- Target Audience: {{ $('Workflow Configuration').item.json.targetAudience }}\n- Desired Tone: {{ $('Workflow Configuration').item.json.tone }}\n\n---\n\nYOUR TASK - Complete these 3 steps:\n\n## STEP 1: Research Trends & Product Perception\nDo both in parallel:\n\n**Trends**: Use Apify to find 5-10 trending TikTok videos. Look for:\n- Viral formats and hooks\n- Popular sounds/music\n- Trending hashtags\n- Content styles that are working right now\n\n**Product Perception**: Use ask_twins to understand how the target audience perceives the product. Include the product name and description in your question. Ask:\n- What problem do they think this solves?\n- What would make them interested?\n- What concerns or objections might they have?\n\n## STEP 2: Write Scripts\nBased on BOTH the twin insights about product perception AND the trends you found, create 3 different TikTok video scripts (15-30 seconds each) that:\n- Address the interests and objections raised by Digital Twins\n- Incorporate trending elements (formats, hooks, styles)\n- Naturally showcase the product\n- Have viral potential\n\nFor each script, define:\n- HOOK (first 2-3 seconds to stop scrolling)\n- SCRIPT (full voiceover/dialogue)\n- VISUAL DIRECTION (what should be shown on screen)\n\n## STEP 3: Test Scripts with Digital Twins\nUse ask_twins to test your scripts with the target audience.\n\nCRITICAL for ask_twins:\n- It only accepts: audience (demographic string) and questions (array of strings)\n- Digital Twins CANNOT see your scripts - you MUST include the full script content INSIDE each question\n- Make 2 calls:\n  1. Hook test: Embed all 3 hooks, ask which would make them stop scrolling\n  2. Authenticity test: Embed full scripts, ask which feels least like an ad\n\n---\n\nOUTPUT FORMAT:\nAfter completing all 3 steps, respond with ONLY raw JSON. No markdown, no code blocks, no explanation - just the JSON object.\n\nIMPORTANT FOR videoPrompt: Write an ACTUAL detailed video generation prompt (100-150 words), NOT placeholders. Example of a GOOD videoPrompt:\n\"Scene 1 (0-3s): Extreme close-up of fingers tapping a phone screen, soft morning window light from the left, shallow depth of field, quiz app interface glowing. Scene 2 (3-6s): Medium shot of a young woman's face, eyes widening with excitement, natural bedroom lighting, subtle smile forming. Scene 3 (6-8s): Pull back to show her jumping up celebrating, confetti overlay, warm golden hour tones. Style: photorealistic, shot on iPhone 15 Pro, authentic UGC aesthetic, smooth transitions. Color grading: warm and vibrant.\"\n\nJSON structure:\n\n{\n  \"trends\": [\"brief summary of trends found\"],\n  \"productInsights\": {\n    \"appeal\": \"What twins found interesting about the product\",\n    \"concerns\": \"Objections or hesitations they mentioned\"\n  },\n  \"scripts\": [\n    {\n      \"id\": 1,\n      \"hook\": \"...\",\n      \"script\": \"...\",\n      \"visualDirection\": \"...\"\n    }\n  ],\n  \"scriptTestFeedback\": {\n    \"hookWinner\": \"Which hook won and why\",\n    \"authenticityWinner\": \"Which script felt most authentic\",\n    \"concerns\": \"Any concerns raised\"\n  },\n  \"bestScript\": {\n    \"id\": 1,\n    \"hook\": \"...\",\n    \"script\": \"...\",\n    \"visualDirection\": \"...\"\n  },\n  \"videoPrompt\": \"Scene 1 (0-3s): [WRITE detailed shot - camera angle, lighting, subject, action]. Scene 2 (3-6s): [WRITE next shot with same detail]. Scene 3 (6-8s): [WRITE final shot]. Style: photorealistic, iPhone 15 Pro footage, natural lighting, shallow DOF, 4K. Color: [warm/cool/vibrant]. THIS MUST BE 100-150 WORDS OF ACTUAL SCENE DESCRIPTIONS, NOT PLACEHOLDERS.\",\n  \"reasoning\": \"Why this script won based on Digital Twin feedback\",\n  \"caption\": \"Catchy TikTok caption under 150 characters that drives engagement\",\n  \"hashtags\": [\"trending1\", \"niche1\", \"product1\", \"viral\", \"fyp\"]\n}\n\nCRITICAL: Output ONLY the JSON object. Do not wrap in ```json``` code blocks. Do not add any text before or after the JSON.",
        "options": {},
        "promptType": "define"
      },
      "typeVersion": 3.1
    },
    {
      "id": "openai-model",
      "name": "OpenAI Chat Model",
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
      "position": [
        992,
        432
      ],
      "parameters": {
        "model": {
          "__rl": true,
          "mode": "list",
          "value": "gpt-4o",
          "cachedResultName": "gpt-4o"
        },
        "options": {}
      },
      "typeVersion": 1.3
    },
    {
      "id": "apify-mcp-tool",
      "name": "Apify MCP",
      "type": "@n8n/n8n-nodes-langchain.mcpClientTool",
      "position": [
        1136,
        432
      ],
      "parameters": {
        "options": {},
        "endpointUrl": "https://mcp.apify.com/sse",
        "authentication": "headerAuth",
        "serverTransport": "sse"
      },
      "typeVersion": 1.2
    },
    {
      "id": "ov-mcp-tool",
      "name": "OriginalVoices Digital Twins",
      "type": "@n8n/n8n-nodes-langchain.mcpClientTool",
      "position": [
        1264,
        432
      ],
      "parameters": {
        "options": {},
        "endpointUrl": "https://api.originalvoices.ai/mcp",
        "authentication": "headerAuth"
      },
      "typeVersion": 1.2
    },
    {
      "id": "parse-json",
      "name": "Parse JSON",
      "type": "n8n-nodes-base.code",
      "position": [
        1280,
        208
      ],
      "parameters": {
        "jsCode": "const output = $input.first().json.output;\n// Strip markdown code blocks if present\nlet jsonStr = output;\nif (typeof output === 'string') {\n  jsonStr = output.replace(/```json\\n?/g, '').replace(/```\\n?/g, '').trim();\n}\nconst parsed = typeof jsonStr === 'string' ? JSON.parse(jsonStr) : jsonStr;\nreturn [{ json: parsed }];"
      },
      "typeVersion": 2
    },
    {
      "id": "generate-video-veo3",
      "name": "Generate Video (Veo 3)",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        1664,
        208
      ],
      "parameters": {
        "url": "https://queue.fal.run/fal-ai/veo3.1",
        "method": "POST",
        "options": {
          "response": {
            "response": {
              "responseFormat": "json"
            }
          }
        },
        "jsonBody": "={\n  \"prompt\": \"{{ $json.videoPrompt }}\",\n  \"aspect_ratio\": \"9:16\",\n  \"duration\": \"8s\"\n}",
        "sendBody": true,
        "specifyBody": "json",
        "authentication": "genericCredentialType",
        "genericAuthType": "httpHeaderAuth"
      },
      "typeVersion": 4.2
    },
    {
      "id": "poll-video-status",
      "name": "Poll Video Status",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        1760,
        432
      ],
      "parameters": {
        "url": "=https://queue.fal.run/fal-ai/veo3.1/requests/{{ $json.request_id }}",
        "options": {
          "response": {
            "response": {
              "responseFormat": "json"
            }
          }
        },
        "authentication": "genericCredentialType",
        "genericAuthType": "httpHeaderAuth"
      },
      "typeVersion": 4.2
    },
    {
      "id": "telegram-confirm",
      "name": "Confirm via Telegram",
      "type": "n8n-nodes-base.telegram",
      "position": [
        2144,
        272
      ],
      "parameters": {
        "text": "=Video ready!\n\nHook:\n{{ $('Parse JSON').item.json.bestScript.hook }}\n\nCaption:\n{{ $('Parse JSON').item.json.caption }} {{ $('Parse JSON').item.json.hashtags.map(h => '#' + h).join(' ') }}\n\nVideo URL:\n{{ $json.video?.url || $json.output?.video?.url || 'Video generation in progress' }}",
        "chatId": "={{ $('Telegram Trigger').item.json.message.chat.id }}",
        "additionalFields": {}
      },
      "typeVersion": 1.2
    },
    {
      "id": "wait-for-video",
      "name": "Wait for Video",
      "type": "n8n-nodes-base.wait",
      "position": [
        1568,
        432
      ],
      "parameters": {
        "amount": 60
      },
      "typeVersion": 1.1
    }
  ],
  "active": false,
  "settings": {
    "executionOrder": "v1"
  },
  "connections": {
    "Apify MCP": {
      "ai_tool": [
        [
          {
            "node": "Research, Write & Test",
            "type": "ai_tool",
            "index": 0
          }
        ]
      ]
    },
    "Parse JSON": {
      "main": [
        [
          {
            "node": "Generate Video (Veo 3)",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Wait for Video": {
      "main": [
        [
          {
            "node": "Poll Video Status",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Telegram Trigger": {
      "main": [
        [
          {
            "node": "Workflow Configuration",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "OpenAI Chat Model": {
      "ai_languageModel": [
        [
          {
            "node": "Research, Write & Test",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "Poll Video Status": {
      "main": [
        [
          {
            "node": "Confirm via Telegram",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Generate Video (Veo 3)": {
      "main": [
        [
          {
            "node": "Wait for Video",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Research, Write & Test": {
      "main": [
        [
          {
            "node": "Parse JSON",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Workflow Configuration": {
      "main": [
        [
          {
            "node": "Research, Write & Test",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "OriginalVoices Digital Twins": {
      "ai_tool": [
        [
          {
            "node": "Research, Write & Test",
            "type": "ai_tool",
            "index": 0
          }
        ]
      ]
    }
  }
}

About this workflow

Create TikTok videos grounded in real human insight. One agent researches viral trends, gathers authentic audience feedback on your product to inform script creation, tests script variations with target audiences to identify the most resonant script option, generates the video,…

Source: https://n8n.io/workflows/13097/ — 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.