Convertisseur de prompt + Veo3

• •Clé API OpenAI

{
  "id": "WCgbgxpfV91txCUx",
  "meta": {
    "instanceId": "c2650793f644091dc80fb900fe63448ad1f4b774008de9608064d67294f8307c",
    "templateCredsSetupCompleted": true
  },
  "name": "Prompt converter + Veo3",
  "tags": [],
  "nodes": [
    {
      "id": "f3ab3272-fe32-4691-bbf2-cfe654b2f7d5",
      "name": "Générer une vidéo",
      "type": "@n8n/n8n-nodes-langchain.googleGemini",
      "position": [
        624,
        0
      ],
      "parameters": {
        "prompt": "={{ JSON.stringify($json.output) }}",
        "modelId": {
          "__rl": true,
          "mode": "list",
          "value": "models/veo-2.0-generate-001",
          "cachedResultName": "models/veo-2.0-generate-001"
        },
        "options": {
          "aspectRatio": "16:9",
          "sampleCount": 1
        },
        "resource": "video"
      },
      "credentials": {
        "googlePalmApi": {
          "id": "7ZkDDnjsiPRXf6eh",
          "name": "project gcp: n8n-khmuhtadin"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "b3227853-e2ba-424b-99b8-b93221f725bb",
      "name": "Analyseur Json",
      "type": "@n8n/n8n-nodes-langchain.outputParserStructured",
      "position": [
        416,
        176
      ],
      "parameters": {
        "jsonSchemaExample": "{\n  \"prompt\": \"detailed scene description\",\n  \"negative_prompt\": \"elements to avoid\",\n  \"aspect_ratio\": \"width:height ratio\",\n  \"camera_movement\": \"camera motion type\",\n  \"style\": \"visual style/genre\",\n  \"quality\": \"output quality setting\",\n  \"motion_intensity\": \"low/medium/high\",\n  \"seed\": \"random seed for reproducibility\",\n  \"fps\": \"frames per second\",\n  \"resolution\": \"video resolution\",\n  \"lighting\": \"lighting conditions\",\n  \"mood\": \"overall mood/atmosphere\",\n  \"color_palette\": \"dominant colors\",\n  \"subject_focus\": \"main subject details\",\n  \"background\": \"background description\",\n  \"effects\": \"special effects to include\"\n}"
      },
      "typeVersion": 1.3
    },
    {
      "id": "c0d14df0-b1eb-4cc3-bd97-be5fffbb4f62",
      "name": "Openai",
      "type": "@n8n/n8n-nodes-langchain.lmChatAzureOpenAi",
      "position": [
        256,
        176
      ],
      "parameters": {
        "model": "gpt",
        "options": {}
      },
      "credentials": {
        "azureOpenAiApi": {
          "id": "IJKdD72VLSmWGj9g",
          "name": "GPT-4.1"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "f5acc743-42d0-4cfa-b202-7d836c57a881",
      "name": "Entrée de prompt",
      "type": "@n8n/n8n-nodes-langchain.chatTrigger",
      "position": [
        0,
        0
      ],
      "webhookId": "a1610267-dc93-4a62-9cf3-d26af6073824",
      "parameters": {
        "options": {}
      },
      "typeVersion": 1.1
    },
    {
      "id": "4cafc8f1-7936-4138-a49c-593ebf47e6f5",
      "name": "Convertisseur de prompt",
      "type": "@n8n/n8n-nodes-langchain.chainLlm",
      "position": [
        224,
        0
      ],
      "parameters": {
        "batching": {},
        "messages": {
          "messageValues": [
            {
              "message": "=# Veo 3 Video Generation JSON Converter System Prompt\n\nYou are a specialized AI assistant that converts natural language descriptions into structured JSON format optimized for Veo 3 video generation. Your primary function is to transform user input into a comprehensive JSON prompt that maximizes the quality and accuracy of generated videos.\n\n## Core Instructions\n\n1. **Always respond in valid JSON format only**\n2. **Never include explanatory text outside the JSON structure**\n3. **Extract and structure all relevant video generation parameters from user input**\n4. **Fill in reasonable defaults for missing parameters**\n5. **Ensure all JSON keys and values follow Veo 3 specifications**\n\n## JSON Structure Template\n\n```json\n{\n  \"prompt\": \"detailed scene description\",\n  \"negative_prompt\": \"elements to avoid\",\n  \"aspect_ratio\": \"width:height ratio\",\n  \"camera_movement\": \"camera motion type\",\n  \"style\": \"visual style/genre\",\n  \"quality\": \"output quality setting\",\n  \"motion_intensity\": \"low/medium/high\",\n  \"seed\": \"random seed for reproducibility\",\n  \"fps\": \"frames per second\",\n  \"resolution\": \"video resolution\",\n  \"lighting\": \"lighting conditions\",\n  \"mood\": \"overall mood/atmosphere\",\n  \"color_palette\": \"dominant colors\",\n  \"subject_focus\": \"main subject details\",\n  \"background\": \"background description\",\n  \"effects\": \"special effects to include\"\n}\n```\n\n## Parameter Guidelines\n\n### Prompt Construction\n- Create detailed, cinematic descriptions\n- Include specific visual elements, actions, and compositions\n- Use descriptive adjectives for style and mood\n- Specify camera angles and movements when relevant\n- Include temporal elements (time of day, season, etc.)\n\n### Default Values\n- **duration**: \"5\" (seconds)\n- **aspect_ratio**: \"16:9\"\n- **quality**: \"high\"\n- **fps**: \"24\"\n- **resolution**: \"1080p\"\n- **motion_intensity**: \"medium\"\n\n### Camera Movements\n- static, pan_left, pan_right, tilt_up, tilt_down, zoom_in, zoom_out, dolly_forward, dolly_backward, orbit, handheld, smooth_tracking\n\n### Style Options\n- cinematic, documentary, animation, vintage, modern, artistic, realistic, stylized, noir, vibrant, minimalist\n\n### Aspect Ratios\n- 16:9 (landscape), 9:16 (portrait), 1:1 (square), 21:9 (ultra-wide), 4:3 (classic)\n\n## Processing Rules\n\n1. **Analyze Input**: Extract key elements like subjects, actions, settings, mood, style preferences\n2. **Enhance Description**: Expand brief descriptions into detailed, cinematic prompts\n3. **Set Parameters**: Choose appropriate technical settings based on content type\n4. **Optimize for Veo 3**: Structure prompt for maximum model comprehension\n5. **Include Negatives**: Add relevant negative prompts to avoid unwanted elements\n\n## Example Transformations\n\n**User Input**: \"A cat playing in the garden\"\n**Output**:\n```json\n{\n  \"prompt\": \"A playful orange tabby cat chasing a butterfly in a sunlit garden, lush green grass, colorful flowers blooming, warm golden hour lighting, shallow depth of field, natural movement\",\n  \"negative_prompt\": \"blurry, low quality, distorted, unnatural movement\",\n  \"aspect_ratio\": \"16:9\",\n  \"camera_movement\": \"smooth_tracking\",\n  \"style\": \"cinematic\",\n  \"quality\": \"high\",\n  \"motion_intensity\": \"medium\",\n  \"fps\": \"24\",\n  \"resolution\": \"1080p\",\n  \"lighting\": \"golden_hour\",\n  \"mood\": \"cheerful\",\n  \"color_palette\": \"warm_natural_tones\"\n}\n```\n\n## Important Notes\n\n- Always output valid JSON only\n- Never explain your reasoning outside the JSON\n- If input is unclear, make reasonable assumptions\n- Prioritize visual storytelling in prompt construction\n- Consider technical limitations of video generation\n- Optimize for realistic and achievable results\n\n## Quality Checklist\n\nBefore outputting JSON, ensure:\n- [ ] Prompt is detailed and descriptive\n- [ ] All technical parameters are appropriate\n- [ ] Negative prompt prevents common issues\n- [ ] JSON syntax is valid\n- [ ] All required fields are present\n- [ ] Values match expected formats"
            }
          ]
        },
        "hasOutputParser": true
      },
      "typeVersion": 1.7
    },
    {
      "id": "ef3a1e96-5bf0-4190-b42a-762fb954ecb2",
      "name": "Alternative",
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
      "position": [
        128,
        176
      ],
      "parameters": {
        "options": {}
      },
      "credentials": {
        "openRouterApi": {
          "id": "DPrHlBN0s5WFDxi9",
          "name": "OpenRouter account"
        }
      },
      "typeVersion": 1
    }
  ],
  "active": false,
  "pinData": {},
  "settings": {
    "executionOrder": "v1"
  },
  "versionId": "ebc10653-9e6e-4b46-b62c-c3d0360997b1",
  "connections": {
    "c0d14df0-b1eb-4cc3-bd97-be5fffbb4f62": {
      "ai_languageModel": [
        [
          {
            "node": "4cafc8f1-7936-4138-a49c-593ebf47e6f5",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "b3227853-e2ba-424b-99b8-b93221f725bb": {
      "ai_outputParser": [
        [
          {
            "node": "4cafc8f1-7936-4138-a49c-593ebf47e6f5",
            "type": "ai_outputParser",
            "index": 0
          }
        ]
      ]
    },
    "f5acc743-42d0-4cfa-b202-7d836c57a881": {
      "main": [
        [
          {
            "node": "4cafc8f1-7936-4138-a49c-593ebf47e6f5",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "4cafc8f1-7936-4138-a49c-593ebf47e6f5": {
      "main": [
        [
          {
            "node": "f3ab3272-fe32-4691-bbf2-cfe654b2f7d5",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}