{
"name": "Product Image Describer powered by easybits",
"tags": [],
"nodes": [
{
"id": "ace9edef-bba9-408d-9f97-d0ff35d49b4c",
"name": "On form submission",
"type": "n8n-nodes-base.formTrigger",
"position": [
190,
377
],
"parameters": {
"options": {},
"formTitle": "Product Image Description Generator",
"formFields": {
"values": [
{
"fieldType": "file",
"fieldLabel": "Product Image",
"requiredField": true,
"acceptFileTypes": ".jpg,.jpeg,.png"
},
{
"fieldLabel": "Product Name"
}
]
},
"formDescription": "Upload a product photo and get a ready-to-use description back."
},
"typeVersion": 2.5
},
{
"id": "506f3f43-9d9f-40e3-b0f8-6869d82cc8cb",
"name": "Code: Expand images into items",
"type": "n8n-nodes-base.code",
"position": [
750,
377
],
"parameters": {
"jsCode": "// Fan out one item per uploaded image, rename its binary to \"data\",\n// and carry a base64 data URI so we can show the image on the result screen\nconst out = [];\nconst inputItems = $input.all();\nlet counter = 0;\n\nfor (let itemIndex = 0; itemIndex < inputItems.length; itemIndex++) {\n const item = inputItems[itemIndex];\n const binaries = item.binary || {};\n const productName = item.json['Product Name'] || item.json.productName || '';\n const keys = Object.keys(binaries);\n\n for (let i = 0; i < keys.length; i++) {\n const key = keys[i];\n const bin = binaries[key];\n counter += 1;\n\n const buffer = await this.helpers.getBinaryDataBuffer(itemIndex, key);\n const mimeType = bin.mimeType || 'image/jpeg';\n const dataUri = `data:${mimeType};base64,${buffer.toString('base64')}`;\n\n out.push({\n json: {\n productName,\n imageIndex: counter,\n fileName: bin.fileName || `image_${counter}`,\n mimeType,\n dataUri,\n },\n binary: { data: bin },\n pairedItem: { item: 0 },\n });\n }\n}\n\nreturn out;"
},
"typeVersion": 2
},
{
"id": "5ee27abc-14d5-496a-9c5d-ecd46f09cb11",
"name": "Loop: Over each image",
"type": "n8n-nodes-base.splitInBatches",
"position": [
1310,
377
],
"parameters": {
"options": {}
},
"typeVersion": 3
},
{
"id": "e3ec54ba-6bb9-4d9d-8d5a-695e7cf335bb",
"name": "Describe: Product image",
"type": "@easybits/n8n-nodes-extractor.easybitsExtractor",
"position": [
1310,
919
],
"parameters": {},
"typeVersion": 2
},
{
"id": "73b8494a-0f4d-478e-9c97-41c58a5445bc",
"name": "Code: Combine descriptions",
"type": "n8n-nodes-base.code",
"position": [
1870,
377
],
"parameters": {
"jsCode": "const results = $input.all();\nconst inputs = $('Code: Expand images into items').all();\n\nconst isMissing = (v) => {\n if (v === null || v === undefined) return true;\n const s = String(v).trim().toLowerCase();\n return s === '' || s === 'null';\n};\nconst esc = (s) => String(s)\n .replace(/&/g,'&').replace(//g,'>').replace(/\"/g,'"');\n\nconst blocks = [];\nresults.forEach((res, i) => {\n const d = res.json.data || res.json || {};\n const meta = (inputs[i] && inputs[i].json) || {};\n\n const fileName = isMissing(meta.fileName) ? `Image ${i+1}` : String(meta.fileName).trim();\n const dataUri = isMissing(meta.dataUri) ? '' : meta.dataUri;\n const rawDesc = isMissing(d.description) ? '' : String(d.description).trim();\n\n const unclear = (rawDesc === '' || rawDesc.toUpperCase() === 'UNCLEAR');\n const body = unclear\n ? 'We could not generate a description for this image. Try a clearer, well-lit photo of a single product.'\n : rawDesc;\n\n const img = dataUri\n ? ``\n : '';\n\n const copyBtn = unclear ? '' :\n ``;\n\n blocks.push(\n `
\n ${img}\n
\n
${esc(fileName)}
\n
${esc(body)}
\n ${copyBtn}\n
\n
`\n );\n});\n\nconst heading = results.length === 1 ? 'Your image description is ready' : 'Your image descriptions are ready';\n\nconst html =\n `
\n
${heading}
\n ${blocks.join('\\n')}\n
`;\n\nreturn [{ json: { imageCount: results.length, combinedDescription: html } }];"
},
"typeVersion": 2
},
{
"id": "6b8eeb52-ad4d-4680-a78c-cae3bd4e3717",
"name": "Show: Result",
"type": "n8n-nodes-base.form",
"position": [
2430,
377
],
"parameters": {
"operation": "completion",
"respondWith": "showText",
"responseText": "={{ $json.combinedDescription }}"
},
"typeVersion": 2.5
},
{
"id": "7402039f-7b7e-426c-8b68-e8a0b1633e78",
"name": "Sticky Note",
"type": "n8n-nodes-base.stickyNote",
"position": [
0,
100
],
"parameters": {
"color": 7,
"width": 480,
"height": 422,
"content": "### Step 1 \u2013 Form intake\nCollects one or more product images (JPG/PNG) and an optional product name. This node's URL is the public form users open."
},
"typeVersion": 1
},
{
"id": "b1e5b542-4494-4525-b9f4-682bbffa469a",
"name": "Sticky Note1",
"type": "n8n-nodes-base.stickyNote",
"position": [
560,
100
],
"parameters": {
"color": 7,
"width": 480,
"height": 422,
"content": "### Step 2 \u2013 Split images into items\nFans each uploaded file out into its own item, renames the binary to `data` for the Extractor, and stores a base64 copy so the image can be shown on the result screen."
},
"typeVersion": 1
},
{
"id": "13f2867d-e89f-4e90-b3c4-c4cee0eef686",
"name": "Sticky Note2",
"type": "n8n-nodes-base.stickyNote",
"position": [
1120,
100
],
"parameters": {
"color": 7,
"width": 480,
"height": 422,
"content": "### Step 3 \u2013 Loop per image\nBatch size **1**, so the Extractor runs once per image instead of bundling them into a single call. The `loop` output feeds the Extractor; the `done` output continues once every image is processed."
},
"typeVersion": 1
},
{
"id": "e05245b4-363b-495d-aeb5-789a2d7bd5ff",
"name": "Sticky Note3",
"type": "n8n-nodes-base.stickyNote",
"position": [
1120,
642
],
"parameters": {
"color": 7,
"width": 480,
"height": 422,
"content": "### Step 4 \u2013 Describe with the easybits Extractor\nRuns each image through the Extractor pipeline and returns structured fields (`product_type`, `description`) under a `data` envelope. Reads the image from the `data` binary property set in Step 2."
},
"typeVersion": 1
},
{
"id": "cddce8e4-6d37-49e3-a376-50046e1fe303",
"name": "Sticky Note4",
"type": "n8n-nodes-base.stickyNote",
"position": [
1680,
100
],
"parameters": {
"color": 7,
"width": 480,
"height": 422,
"content": "### Step 5 \u2013 Build the result\nReads every per-image result from the loop's done branch, pairs each with its filename and thumbnail, and builds the HTML for the result page. Images the model can't read come back as `UNCLEAR` and get a fallback message."
},
"typeVersion": 1
},
{
"id": "4e2d121e-9d6f-4ea1-a199-6344e99ad46d",
"name": "Sticky Note5",
"type": "n8n-nodes-base.stickyNote",
"position": [
2240,
100
],
"parameters": {
"color": 7,
"width": 480,
"height": 422,
"content": "### Step 6 \u2013 Show the result\nDisplays the descriptions in the browser with a copy button per image, using the Form completion screen (Show Text renders the HTML)."
},
"typeVersion": 1
},
{
"id": "f7de45f9-a155-4184-a3af-8d2ce58ffa31",
"name": "Sticky Note6",
"type": "n8n-nodes-base.stickyNote",
"position": [
-764,
100
],
"parameters": {
"width": 704,
"height": 1068,
"content": "## \ud83d\uddbc\ufe0f Product Image Describer (powered by easybits)\n\nUpload one or more product photos and get back a clean, ready-to-use\ndescription for each \u2013 with a copy button to paste straight into your shop.\n\n## How it works\n1. **Form intake** \u2013 the user uploads product images (JPG/PNG) and an optional product name.\n2. **Split** \u2013 each image becomes its own item and is prepped for the Extractor.\n3. **Loop** \u2013 images are processed one at a time, so each gets its own description.\n4. **Describe** \u2013 the easybits Extractor reads each image and returns structured fields (product type + description).\n5. **Build & show** \u2013 results are combined into a styled page with a thumbnail, filename and copy button per image.\n\n## Set up steps\n1. **Add the Extractor node** (`Describe: Product image`):\n - **n8n Cloud:** verified node \u2013 just search \"easybits Extractor\" in the node panel. Nothing to install.\n - **Self-hosted:** Settings \u2192 Community Nodes \u2192 Install \u2192 `@easybits/n8n-nodes-extractor`\n2. **Create an Extractor pipeline** at extractor.easybits.tech with two String fields:\n - `product_type`\n - `description`\n Then paste the Pipeline ID + API key into the node's credential.\n3. The Extractor **free plan includes 50 API requests / month** \u2013 enough to test this end to end.\n4. **Activate** the workflow and open the form URL to try it.\n\n\n**Good to know:** in the `description` field, tell the model not to guess at\nbrand, price or size, and to return `UNCLEAR` if it can't read the image \u2013\nthe workflow shows a friendly fallback line for those. ~2 seconds per image."
},
"typeVersion": 1
}
],
"active": false,
"settings": {
"availableInMCP": false,
"executionOrder": "v1"
},
"connections": {
"On form submission": {
"main": [
[
{
"node": "Code: Expand images into items",
"type": "main",
"index": 0
}
]
]
},
"Loop: Over each image": {
"main": [
[
{
"node": "Code: Combine descriptions",
"type": "main",
"index": 0
}
],
[
{
"node": "Describe: Product image",
"type": "main",
"index": 0
}
]
]
},
"Describe: Product image": {
"main": [
[
{
"node": "Loop: Over each image",
"type": "main",
"index": 0
}
]
]
},
"Code: Combine descriptions": {
"main": [
[
{
"node": "Show: Result",
"type": "main",
"index": 0
}
]
]
},
"Code: Expand images into items": {
"main": [
[
{
"node": "Loop: Over each image",
"type": "main",
"index": 0
}
]
]
}
}
}