Vasyl Pavlyuchok This workflow corresponds to n8n.io template #10519 — we link there as the canonical source.
This workflow follows the Googlegemini → HTTP Request 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 →
{
"meta": {
"templateCredsSetupCompleted": true
},
"nodes": [
{
"id": "43e712ff-dbda-4924-8d45-faac98713e20",
"name": "Sticky Note20",
"type": "n8n-nodes-base.stickyNote",
"position": [
416,
240
],
"parameters": {
"width": 272,
"height": 352,
"content": "## 3. Result Splitter\nSplitting feed results to separe items"
},
"typeVersion": 1
},
{
"id": "b927d71f-0547-4051-afeb-f00e5ce16241",
"name": "Sticky Note21",
"type": "n8n-nodes-base.stickyNote",
"position": [
720,
240
],
"parameters": {
"color": 7,
"width": 320,
"height": 352,
"content": "## 3. Filter by last 24h\nYou can edit HOURS in the code if you need a wider window."
},
"typeVersion": 1
},
{
"id": "4d5cd7cf-0c31-45a4-838a-92505b192056",
"name": "Sticky Note22",
"type": "n8n-nodes-base.stickyNote",
"position": [
1552,
240
],
"parameters": {
"color": 4,
"width": 880,
"height": 352,
"content": "## 5. Register in Notion\n* Filter fields before Notion\n* Register to Notion\n* Freeze **page_id** of each record\n* Add subtitle **Deep Research Summary** inside new pages (will be used later)"
},
"typeVersion": 1
},
{
"id": "f3a392a2-2e79-4c81-89e7-f19f749b0e86",
"name": "Sticky Note23",
"type": "n8n-nodes-base.stickyNote",
"position": [
3968,
464
],
"parameters": {
"color": 4,
"width": 256,
"height": 352,
"content": "## 10. Saving to Notion\nEach chunk is saved, one by one, under the related paper according to their **page_id**"
},
"typeVersion": 1
},
{
"id": "7c7b1d33-ce9b-49fe-b94f-a1a2daa56ba9",
"name": "Sticky Note24",
"type": "n8n-nodes-base.stickyNote",
"position": [
2464,
240
],
"parameters": {
"color": 7,
"width": 2240,
"height": 800,
"content": "## 7. Generating and Saving a Deep Research Summary for each Paper\nProcessing every paper (full documents), one by one. Then generating a **Deep Research Summary** for each document and saving it to Notion. The **Deep Research Summary** provides a deeper analysis explaining paper's ideas and meanings clearly to curious professionals and non technical experts.\n\nContents Generated:\n-> 1) 3\u20135 core insights; 2) Core idea and motivation \u2014 explained simply; 3) Why this research matters now (context & importance); 4) Key innovations & contributions; 5) Method \u2014 explained step-by-step in plain language; 6) Math / theory intuition (no formulas); 7) Experiments and evaluation \u2014 what was tested and why it matters; 8) Key results & what they prove; 9) Limitations / where it may fail; 10) Real-world impact and applications; 11) Future work / open questions; 12) Closing takeaway \u2014 2-3 sentences summarizing the big picture"
},
"typeVersion": 1
},
{
"id": "edafcf5f-5cf6-409a-af6e-a75a4149e6f9",
"name": "Sticky Note25",
"type": "n8n-nodes-base.stickyNote",
"position": [
2224,
-96
],
"parameters": {
"color": 5,
"width": 528,
"height": 304,
"content": "## 6. Send News to Telegram\nMessages are sent as separate items, which include **title**, **authors**, **data** and **links** (including a link to Notion with **Deep Resume** we will be generating)."
},
"typeVersion": 1
},
{
"id": "003c9df5-db71-4921-a212-39048bac0c17",
"name": "Sticky Note26",
"type": "n8n-nodes-base.stickyNote",
"position": [
1072,
240
],
"parameters": {
"width": 448,
"height": 352,
"content": "## 4. Result Cleaner\nDedupe results in case of repeated items and prepare clean fields for the next node"
},
"typeVersion": 1
},
{
"id": "91ca89b2-1c87-4baa-a490-d10b3bc73bdb",
"name": "Sticky Note28",
"type": "n8n-nodes-base.stickyNote",
"position": [
-288,
-240
],
"parameters": {
"color": 3,
"width": 4992,
"height": 80,
"content": "# AI NEWS FROM ARXIV"
},
"typeVersion": 1
},
{
"id": "dbd106b3-5991-4279-b6a1-6ac708e20242",
"name": "Sticky Note29",
"type": "n8n-nodes-base.stickyNote",
"position": [
3440,
464
],
"parameters": {
"width": 496,
"height": 352,
"content": "## 9. Splitting Output\nThe Output Data is too long to be saved to Notion, so is splitted into different chunks [<2000 characters], then debuged before saving"
},
"typeVersion": 1
},
{
"id": "4a019d5e-d657-476e-80f8-2f4dd0ef03e5",
"name": "Sticky Note30",
"type": "n8n-nodes-base.stickyNote",
"position": [
2832,
464
],
"parameters": {
"color": 3,
"width": 576,
"height": 464,
"content": "## 8. Generating Deep Research Summary\nEach paper's pfd is analized to get a **Deep Research Summary**, and then **merge** the result with the related **page_id** to be saved to Notion."
},
"typeVersion": 1
},
{
"id": "65ba3fb9-7d0d-4a66-b658-5d549366f188",
"name": "Code (Parse Gemini JSON in chunks)7",
"type": "n8n-nodes-base.code",
"position": [
3552,
608
],
"parameters": {
"jsCode": "// Toma todos los items del Merge\nconst items = $input.all();\n\n// 1) Localiza page_id y el texto de Gemini\nconst pageId =\n items.find(it => it.json?.page_id)?.json.page_id ?? null;\n\nconst rawText =\n items.find(it => it.json?.content)?.json?.content?.parts?.[0]?.text ?? \"\";\n\n// 2) Limpia fences ```json ... ``` y espacios\nconst cleaned = String(rawText)\n .replace(/```json/gi, \"\")\n .replace(/```/g, \"\")\n .trim();\n\n// 3) Intenta parsear JSON. Si falla, deja objeto vac\u00edo.\nlet parsed = {};\ntry {\n parsed = cleaned ? JSON.parse(cleaned) : {};\n} catch (_) {\n parsed = {};\n}\n\n// 4) Extrae el array de trozos con tolerancia de nombres\nconst chunks =\n Array.isArray(parsed.chunks) ? parsed.chunks :\n Array.isArray(parsed.summary_chunks) ? parsed.summary_chunks :\n Array.isArray(parsed.slices) ? parsed.slices :\n [];\n\n// 5) Si no hay page_id o no hay chunks, no emites nada \u00fatil\nif (!pageId || !Array.isArray(chunks) || chunks.length === 0) {\n // Opcional: emite un item de depuraci\u00f3n para ver qu\u00e9 lleg\u00f3\n return [{\n json: {\n __parse_debug: true,\n page_id_present: !!pageId,\n got_chunks: Array.isArray(chunks) ? chunks.length : -1,\n preview: cleaned.slice(0, 280)\n }\n }];\n}\n\n// 6) Emite un item por chunk: { page_id, slice, chunk_index, slice_index }\nreturn chunks.map((chunk, i) => ({\n json: {\n page_id: pageId,\n slice: String(chunk).trim(),\n chunk_index: i,\n slice_index: 0,\n }\n}));\n"
},
"typeVersion": 2
},
{
"id": "fedd41d3-62ac-4685-a66f-8c0e5a42a329",
"name": "Split Results",
"type": "n8n-nodes-base.splitOut",
"position": [
496,
400
],
"parameters": {
"options": {},
"fieldToSplitOut": "feed.entry"
},
"typeVersion": 1
},
{
"id": "86faed42-ea10-499f-bbbf-c2edca9dffd8",
"name": "Sticky Note32",
"type": "n8n-nodes-base.stickyNote",
"position": [
16,
240
],
"parameters": {
"width": 368,
"height": 352,
"content": "## 2. Website Scraper\nScraping papers from [arXiv.org] filtered by **Artificial Intelligence** and then converting the results to JSON"
},
"typeVersion": 1
},
{
"id": "a3b34013-d1e8-429e-a998-704bcb3bee13",
"name": "Sticky Note33",
"type": "n8n-nodes-base.stickyNote",
"position": [
-288,
240
],
"parameters": {
"color": 6,
"width": 272,
"height": 368,
"content": "## 1. Daily arXiv AI News Trigger\nExecuted daily at 8.00 am CET\n\nlink: https://arxiv.org/search/?query=Artificial+Intelligence&searchtype=all&source=header"
},
"typeVersion": 1
},
{
"id": "95d19cfc-f7d9-4326-bda7-2deaebaedfdd",
"name": "Scheduled Daily Trigger",
"type": "n8n-nodes-base.scheduleTrigger",
"position": [
-208,
448
],
"parameters": {
"rule": {
"interval": [
{
"triggerAtHour": 8
}
]
}
},
"typeVersion": 1.2
},
{
"id": "befe2fc3-f00e-4157-a4ba-fbdbb4a003dc",
"name": "HTTP Request",
"type": "n8n-nodes-base.httpRequest",
"position": [
80,
400
],
"parameters": {
"url": "https://export.arxiv.org/api/query",
"options": {
"response": {
"response": {
"responseFormat": "text"
}
},
"allowUnauthorizedCerts": true
},
"sendQuery": true,
"queryParameters": {
"parameters": [
{
"name": "search_query",
"value": "abs:\"artificial intelligence\""
},
{
"name": "sortBy",
"value": "submittedDate"
},
{
"name": "sortOrder",
"value": "descending"
},
{
"name": "start",
"value": "0"
},
{
"name": "max_results",
"value": "100"
}
]
}
},
"typeVersion": 4.2
},
{
"id": "81ea7bd5-edac-43f7-9736-3f5fed6678cc",
"name": "Format Conversor",
"type": "n8n-nodes-base.xml",
"position": [
240,
400
],
"parameters": {
"options": {
"mergeAttrs": true
}
},
"typeVersion": 1
},
{
"id": "c4c53a6d-cfae-4817-b077-5c82f8381fff",
"name": "Dedupe with Static Data",
"type": "n8n-nodes-base.code",
"position": [
1136,
400
],
"parameters": {
"jsCode": "// === DEDUPLICADOR PERSISTENTE ===\n// Guarda IDs ya vistos entre ejecuciones (funciona en \"Code\" con modo Run Once for All Items)\n\nconst store = $getWorkflowStaticData('global'); // \u2705 Correcta sintaxis en n8n actual\nif (!store.seen) store.seen = {}; // mapa id -> fecha\nconst seen = store.seen;\n\n// Dedupe dentro de la misma ejecuci\u00f3n\nconst batchSeen = new Set();\nconst out = [];\n\nfor (const it of items) {\n const id = it.json.arxiv_id || it.json.link || it.json.id;\n if (!id) continue;\n\n if (batchSeen.has(id)) continue; // ya apareci\u00f3 en esta ejecuci\u00f3n\n batchSeen.add(id);\n\n if (seen[id]) continue; // ya apareci\u00f3 en ejecuciones previas\n\n // marcar como nuevo\n seen[id] = new Date().toISOString();\n out.push(it);\n}\n\n// Limpiar registros viejos si excede 2000 entradas\nconst MAX = 2000;\nconst entries = Object.entries(seen)\n .sort((a, b) => new Date(b[1]) - new Date(a[1]));\nif (entries.length > MAX) {\n const keep = new Set(entries.slice(0, MAX).map(([k]) => k));\n for (const k of Object.keys(seen)) {\n if (!keep.has(k)) delete seen[k];\n }\n}\n\nreturn out;\n"
},
"typeVersion": 2
},
{
"id": "d66eb078-88f1-4de0-b3ae-95ad34398c27",
"name": "Edit Fields for Notion (incl. pdf)",
"type": "n8n-nodes-base.set",
"position": [
1632,
400
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "4f50a382-fd77-425b-b583-5140078f1d15",
"name": "title",
"type": "string",
"value": "={{ $json.properties.title.title[0].text.content }}"
},
{
"id": "943afbe5-13a8-46e8-8904-6c9139e741a5",
"name": "published",
"type": "string",
"value": "={{ $json.properties.published.date.start }}"
},
{
"id": "2476f04e-1db7-41aa-9388-bc9d692883bf",
"name": "arxiv_id",
"type": "string",
"value": "={{ $json.properties.arxiv_id.rich_text[0].text.content }}"
},
{
"id": "ca36a055-966b-4b29-a164-cfee360fc999",
"name": "authors",
"type": "string",
"value": "={{ $json.properties.authors.rich_text[0].text.content }}"
},
{
"id": "e820d5fd-d900-4b00-82d3-773a9eafe6a6",
"name": "abstract",
"type": "string",
"value": "={{ $json.properties.abstract.rich_text[0].text.content }}"
},
{
"id": "1e5e2d1b-ae9b-4a7a-b6c3-aa4b80751a61",
"name": "abstract_clean",
"type": "string",
"value": "={{ \n String($json.properties.abstract.rich_text[0].text.content || $json.summary)\n .replace(/\\\\\\\\n/g, ' ') // limpia \"\\n\" escapado\n .replace(/\\\\n/g, ' ') // limpia \"\\n\" si llega doble escapado\n .replace(/\\n/g, ' ') // limpia saltos reales\n .replace(/\\s+/g, ' ') // colapsa espacios m\u00faltiples\n .trim() \n}}"
},
{
"id": "eece341f-e5cc-4d2f-8c0b-66796d7e5c45",
"name": "url_pdf",
"type": "string",
"value": "={{ (() => {\n const src = ($json.link || $json.properties.arxiv_id.rich_text[0].text.content || '').trim();\n if (!src) return '';\n // Cambia /abs/ por /pdf/\n let u = src.replace('/abs/', '/pdf/');\n // Asegura protocolo\n if (!/^https?:\\/\\//i.test(u)) u = 'https://' + u.replace(/^\\/+/, '');\n // A\u00f1ade .pdf si no lo tiene\n if (!u.toLowerCase().endsWith('.pdf')) u += '.pdf';\n return u;\n})() }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "df174b58-d4a9-4bcf-b35d-ef564521408d",
"name": "Register to Notion Database",
"type": "n8n-nodes-base.notion",
"position": [
1840,
400
],
"parameters": {
"title": "={{ $json.title }}",
"options": {},
"resource": "databasePage",
"databaseId": {
"__rl": true,
"mode": "list",
"value": "2992729c-bda2-8008-950c-f617082228a1",
"cachedResultUrl": "https://www.notion.so/2992729cbda28008950cf617082228a1",
"cachedResultName": "ARXIV"
},
"propertiesUi": {
"propertyValues": [
{
"key": "arxiv_id|url",
"urlValue": "={{ $json.arxiv_id }}"
},
{
"key": "authors|rich_text",
"textContent": "={{ $json.authors }}"
},
{
"key": "published|date",
"date": "={{ $json.published }}",
"timezone": "US/Eastern"
},
{
"key": "abstract|rich_text",
"textContent": "={{ $json.abstract_clean }}"
},
{
"key": "url_pdf|url",
"urlValue": "={{ $json.url_pdf }}"
}
]
}
},
"typeVersion": 2.2
},
{
"id": "2e7d8869-f34c-4ba8-a1f3-4e4721cf6916",
"name": "Freeze page_id",
"type": "n8n-nodes-base.set",
"position": [
2048,
400
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "2b4c8c3c-d029-4d03-a74b-85c361e7c0ec",
"name": "=page_id",
"type": "string",
"value": "={{ $json.id }}"
}
]
},
"includeOtherFields": true
},
"typeVersion": 3.4
},
{
"id": "78861c59-f436-4959-9f77-2c919adf512d",
"name": "Append a block (adding \"block\" subtitle)",
"type": "n8n-nodes-base.notion",
"position": [
2256,
400
],
"parameters": {
"blockId": {
"__rl": true,
"mode": "id",
"value": "={{ $json.id }}"
},
"blockUi": {
"blockValues": [
{
"type": "heading_2",
"textContent": "Deep Research Summary"
}
]
},
"resource": "block"
},
"typeVersion": 2.2
},
{
"id": "d20e2281-2f80-4ac9-ab46-fa170379c4f8",
"name": "Send a text message",
"type": "n8n-nodes-base.telegram",
"position": [
2320,
32
],
"parameters": {
"text": "=<b>{{ $('Register to Notion Database').item.json.property_title }}</b>\n\n<b>{{ $('Register to Notion Database').item.json.property_published.start }}</b>\n\n<b>{{ $('Register to Notion Database').item.json.property_authors }}</b>\n\n<em>ABSTRACT (short):</em>\n{{ $('Edit Fields for Notion (incl. pdf)').item.json.abstract_clean }}\n\n<b>Links</b>\n\u2022 arXiv: {{ $json.url || $json.arxiv_id || '' }}\n\u2022 Complete paper (pdf): {{ $json.property_url_pdf || '' }}\n\u2022 Deep Summary (Notion): {{ $('Register to Notion Database').item.json.url }}\n",
"chatId": "=-1003277954309",
"additionalFields": {
"parse_mode": "=HTML",
"appendAttribution": false
}
},
"credentials": {
"telegramApi": {
"name": "<your credential>"
}
},
"typeVersion": 1.2
},
{
"id": "ab41032c-3654-4ea5-83b3-3a690a597ed3",
"name": "Clean page_id",
"type": "n8n-nodes-base.set",
"position": [
2880,
592
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "d928dfb2-65a2-449e-8297-d458471a5075",
"name": "=page_id",
"type": "string",
"value": "={{$json.page_id || $json.parent?.page_id || $json.page_id || $json.results?.[0]?.parent?.page_id || $json.id}}"
}
]
},
"includeOtherFields": true
},
"typeVersion": 3.4
},
{
"id": "276f2649-c421-4913-86e0-3789d7187a24",
"name": "Analyze doc (Prompt Ultra-Pro)",
"type": "@n8n/n8n-nodes-langchain.googleGemini",
"maxTries": 2,
"position": [
3072,
720
],
"parameters": {
"text": "You are an elite AI research analyst and science communicator.\n\nYour task:\nDeeply read and understand the attached research PDF and generate a structured, high-value explainer that will be stored inside a Notion child page.\n\nAudience:\n- Curious, intelligent founders and professionals\n- Not technical experts, but can understand advanced concepts when clearly explained\n\nPrimary mission:\nExplain the paper\u2019s ideas, meaning, and significance \u2014 NOT the metadata.\nDo NOT repeat title, authors, arXiv ID, year, or publication info. We already store those separately.\n\nTone:\n- Clear, sharp, human, insightful\n- Like Karpathy + Paul Graham + Ali Abdaal\n- No fluff, no hype, no academic filler\n- Metaphors and intuitive explanations welcome\n- Convey meaning and understanding, not jargon\n\nCONTENT YOU MUST PRODUCE (as narrative with brief bold section headers inside text):\n\n1) 3\u20135 core insights (executive brief)\n2) Core idea and motivation \u2014 explained simply\n3) Why this research matters now (context & importance)\n4) Key innovations & contributions\n5) Method \u2014 explained step-by-step in plain language\n6) Math / theory intuition (no formulas \u2014 explain what they *mean*)\n7) Experiments and evaluation \u2014 what was tested and why it matters\n8) Key results & what they prove (plain English significance)\n9) Limitations / where it may fail\n10) Real-world impact and applications\n11) Future work / open questions\n12) Closing takeaway \u2014 2-3 sentences summarizing the big picture\n\nOutput rules:\n- DO NOT repeat metadata (title, authors, arXiv ID, etc.)\n- DO NOT output anything outside JSON\n- NO chunk > 1900 characters\n- NO markdown blocks, no backticks inside chunks\n- Make the text feel like a human expert teaching\n\nReturn only valid JSON in this format:\n\n{\n \"chunks\": [\n \"chunk1 <= 1900 chars\",\n \"chunk2 <= 1900 chars\",\n \"...if needed\"\n ]\n}\n\nBegin your analysis now.\n",
"modelId": {
"__rl": true,
"mode": "list",
"value": "models/gemini-2.5-pro",
"cachedResultName": "models/gemini-2.5-pro"
},
"options": {},
"resource": "document",
"documentUrls": "={{ $('Edit Fields for Notion (incl. pdf)').item.json.url_pdf }}"
},
"credentials": {
"googlePalmApi": {
"name": "<your credential>"
}
},
"retryOnFail": true,
"typeVersion": 1
},
{
"id": "86715411-218c-4910-99de-159a0270c1b1",
"name": "Merge page_id & Summary",
"type": "n8n-nodes-base.merge",
"position": [
3264,
608
],
"parameters": {
"mode": "combine",
"options": {},
"combineBy": "combineByPosition"
},
"typeVersion": 3.2
},
{
"id": "89127dd3-5486-430c-9677-e44f36269460",
"name": "Debug before append",
"type": "n8n-nodes-base.set",
"position": [
3744,
608
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "a1465705-0160-46c2-8d10-50beede0eb11",
"name": "page_id",
"type": "string",
"value": "={{$json.db_page_id || $json.page_id}}"
},
{
"id": "9c3f2c3c-1139-4919-a45b-46db0f62659d",
"name": "slice",
"type": "string",
"value": "={{$json.slice}}"
},
{
"id": "f9db996a-ae16-4f8c-89ac-02f7bcc9dcee",
"name": "=dbg_page_id",
"type": "string",
"value": "={{$json.page_id}}"
},
{
"id": "ab9d751f-bd50-45d5-a560-65a8b8650ff1",
"name": "dbg_slice_len",
"type": "string",
"value": "={{$json.slice.length}}"
},
{
"id": "50a5acfd-0878-4340-8ba6-c996b5132dc6",
"name": "dbg_chunk",
"type": "string",
"value": "={{$json.chunk_index}}"
},
{
"id": "0c89088e-fd2b-4286-a1c3-f737efde3ae0",
"name": "dbg_slice_index",
"type": "string",
"value": "={{$json.slice_index}}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "f12f15de-1c28-431b-8102-a54da0541b80",
"name": "Append chunks as blocks",
"type": "n8n-nodes-base.notion",
"onError": "continueRegularOutput",
"position": [
4048,
608
],
"parameters": {
"blockId": {
"__rl": true,
"mode": "id",
"value": "={{$json.page_id}}"
},
"blockUi": {
"blockValues": [
{
"textContent": "={{$json.slice || '\u2014'}}\n"
}
]
},
"resource": "block"
},
"typeVersion": 2.2
},
{
"id": "710ef57c-cb3c-460b-a957-7aae51ddd704",
"name": "Sticky Note27",
"type": "n8n-nodes-base.stickyNote",
"position": [
4256,
672
],
"parameters": {
"width": 400,
"height": 320,
"content": "## 11. Wait and Start Over\nAfter saving a paper's **Deep Research Summary**, wait and start over with remaining ones."
},
"typeVersion": 1
},
{
"id": "f3f8e985-9250-499f-a0e3-8b41bb94b55f",
"name": "When clicking \u2018Execute workflow\u2019",
"type": "n8n-nodes-base.manualTrigger",
"position": [
-208,
752
],
"parameters": {},
"typeVersion": 1
},
{
"id": "eb2b5824-d60b-4e7a-94da-1e16ac3e8514",
"name": "Sticky Note34",
"type": "n8n-nodes-base.stickyNote",
"position": [
-288,
640
],
"parameters": {
"color": 6,
"width": 272,
"height": 304,
"content": "## Manual Trigger\nUse this trigger to test the workflow"
},
"typeVersion": 1
},
{
"id": "897206e4-e3e8-4a92-8592-e90fc1a7d648",
"name": "Sticky Note31",
"type": "n8n-nodes-base.stickyNote",
"position": [
-272,
1104
],
"parameters": {
"color": 3,
"width": 4992,
"height": 80,
"content": ""
},
"typeVersion": 1
},
{
"id": "9bba5011-5203-4b14-b456-d4b1fc46d202",
"name": "Sticky Note11",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1072,
-160
],
"parameters": {
"width": 652,
"height": 1248,
"content": "## Try It Out\n\n### What this template does (end-to-end)\n* Fetches new AI papers from arXiv.\n* Filters by last 24h time window and de-duplicates papers.\n* Creates a Notion database page per paper and stores key fields (+ the PDF link).\n* Generates a plain-English **Deep Research** summary from the PDF (Gemini) and saves it as child blocks in the same Notion page.\n* Posts a Telegram update at **08:00** with title + link(s) + a short abstract, including a link to the Deep Research in Notion.\n\n### Run it now\n* Click **Execute workflow** to test manually.\n* When ready, enable the **Scheduled Trigger** (08:00 local by default).\n\n### Configure these nodes (minimal)\n* **HTTP Request** \u2192 arXiv feed/query (e.g., your AI categories or keywords).\n* **Filter by last 24h** -> you can edit HOURS in the code if you need a wider window.\n* **Result Cleaner** \u2192 keeps first occurrence of each `arxiv_id`.\n* **Register in Notion**\n * Notion account (integration added to the target DB)\n * Database ID and property mapping (`name`, `url`, `abstract`, `authors`, `tags`, `pdf_url`, `arxiv_id`, `status`).\n* **Freeze page_id** \u2192 carries the created Notion `page_id` forward.\n* **Analyze doc (Gemini)** \u2192 set your model + API key (env var recommended).\n* **Code (Parse Gemini JSON in chunks)** \u2192 splits the model output into \u22641900-char slices.\n* **Append chunks (Notion)** \u2192 appends each slice under the same `page_id`.\n* **Send to Telegram** \u2192 bot token + `chat_id` (channel or group).\n\n### What you\u2019ll see\n* One Notion page per paper with metadata, PDF link, and a **Deep Research** section composed of multiple rich-text blocks.\n* One Telegram post per paper with the Notion page link and the PDF.\n\n### Notes & limits\n* The summarizer reads the PDF and returns a structured explainer (core insights, method, results, limits, applications).\n* Chunking ensures each block \u22641900 chars to avoid Notion/API errors.\n* If you expect many items, keep **Wait** enabled to pace API calls.\n\n### Requirements\n* Notion integration with access to your target database.\n* Telegram bot + channel/group `chat_id`.\n* Gemini API key and model (e.g., `gemini-2.5-pro`).\n* Optional: environment variables for secrets and IDs.\n\n### Troubleshooting\n* Empty summaries \u2192 check PDF URL reachability and Gemini key/model.\n* \"Page not found\" on append \u2192 verify `page_id` is set before Append.\n* Duplicates \u2192 confirm `arxiv_id` exists and the dedupe node is before Notion create.\n* Telegram not posting \u2192 bot must be **admin** of the channel.\n"
},
"typeVersion": 1
},
{
"id": "d80aa1ea-5332-46bb-a03c-89a5655039aa",
"name": "End Telegram branch (no further actions)",
"type": "n8n-nodes-base.noOp",
"position": [
2544,
32
],
"parameters": {},
"typeVersion": 1
},
{
"id": "84efcd13-b37e-47fd-95f1-385181986bb3",
"name": "Process each paper (Gemini + Notion loop)",
"type": "n8n-nodes-base.splitInBatches",
"position": [
2592,
480
],
"parameters": {
"options": {}
},
"typeVersion": 3
},
{
"id": "88d53b3c-e277-4cc9-91f9-bc083e41314c",
"name": "Return to paper loop (next paper)",
"type": "n8n-nodes-base.noOp",
"position": [
4496,
784
],
"parameters": {},
"typeVersion": 1
},
{
"id": "1abceb48-d007-4eaf-83ab-8234ecd6db56",
"name": "Delay between paper summaries",
"type": "n8n-nodes-base.wait",
"position": [
4320,
784
],
"parameters": {
"amount": 0.4
},
"typeVersion": 1.1
},
{
"id": "9a6e800c-1fd2-45c3-ac36-2cfb95c259bc",
"name": "Prepare Notion payload (JSON)",
"type": "n8n-nodes-base.code",
"position": [
1360,
400
],
"parameters": {
"jsCode": "return items.map(item => ({\n json: {\n parent: { database_id: \"TU_DATABASE_ID\" },\n properties: {\n title: {\n title: [{ text: { content: item.json.title || \"Sin t\u00edtulo\" } }]\n },\n arxiv_id: {\n rich_text: [{ text: { content: item.json.arxiv_id || \"\" } }]\n },\n abstract: {\n rich_text: [{ text: { content: item.json.summary || \"\" } }]\n },\n authors: {\n rich_text: [{ text: { content: item.json.authors || \"\" } }]\n },\n categories: {\n rich_text: [{ text: { content: item.json.categories || \"\" } }]\n },\n published: {\n date: { start: item.json.published || null }\n },\n link: {\n url: item.json.link || null\n }\n }\n }\n}));\n"
},
"typeVersion": 2
},
{
"id": "34727c62-2923-4868-82c1-037fb449615f",
"name": "Filter recent papers (last 24h)",
"type": "n8n-nodes-base.code",
"position": [
832,
400
],
"parameters": {
"jsCode": "// === Configuraci\u00f3n ===\nconst HOURS = 24; // Cambia a 48 o 72 si quieres ampliar el rango\nconst now = new Date();\nconst threshold = new Date(now.getTime() - HOURS * 60 * 60 * 1000);\n\n// === Utilidad para limpiar texto ===\nconst get = v => Array.isArray(v) ? String(v[0] ?? '').trim() : String(v ?? '').trim();\n\n// === Procesar cada paper ===\nconst out = [];\n\nfor (const it of items) {\n const e = it.json;\n\n // Convertimos fechas\n const published = new Date(get(e.published || e.updated));\n if (isNaN(published) || published < threshold) continue; // filtra los m\u00e1s antiguos\n\n // --- Autores ---\n const authorList = Array.isArray(e.author) ? e.author : [e.author];\n const authors = authorList\n .filter(Boolean)\n .map(a => get(a.name))\n .join(', ');\n\n // --- Categor\u00edas ---\n const categoryList = Array.isArray(e.category) ? e.category : [e.category];\n const categories = categoryList\n .filter(Boolean)\n .map(c => c?.['@_term'] || c?.['$'] || '')\n .filter(Boolean)\n .join(', ');\n\n // --- A\u00f1adir al resultado ---\n out.push({\n json: {\n arxiv_id: get(e.id),\n title: get(e.title),\n summary: get(e.summary),\n authors,\n categories,\n published: published.toISOString(),\n link: get(e.id)\n }\n });\n}\n\nreturn out;\n"
},
"typeVersion": 2
}
],
"connections": {
"HTTP Request": {
"main": [
[
{
"node": "Format Conversor",
"type": "main",
"index": 0
}
]
]
},
"Clean page_id": {
"main": [
[
{
"node": "Analyze doc (Prompt Ultra-Pro)",
"type": "main",
"index": 0
},
{
"node": "Merge page_id & Summary",
"type": "main",
"index": 0
}
]
]
},
"Split Results": {
"main": [
[
{
"node": "Filter recent papers (last 24h)",
"type": "main",
"index": 0
}
]
]
},
"Freeze page_id": {
"main": [
[
{
"node": "Append a block (adding \"block\" subtitle)",
"type": "main",
"index": 0
}
]
]
},
"Format Conversor": {
"main": [
[
{
"node": "Split Results",
"type": "main",
"index": 0
}
]
]
},
"Debug before append": {
"main": [
[
{
"node": "Append chunks as blocks",
"type": "main",
"index": 0
}
]
]
},
"Send a text message": {
"main": [
[
{
"node": "End Telegram branch (no further actions)",
"type": "main",
"index": 0
}
]
]
},
"Append chunks as blocks": {
"main": [
[
{
"node": "Delay between paper summaries",
"type": "main",
"index": 0
}
]
]
},
"Dedupe with Static Data": {
"main": [
[
{
"node": "Prepare Notion payload (JSON)",
"type": "main",
"index": 0
}
]
]
},
"Merge page_id & Summary": {
"main": [
[
{
"node": "Code (Parse Gemini JSON in chunks)7",
"type": "main",
"index": 0
}
]
]
},
"Scheduled Daily Trigger": {
"main": [
[
{
"node": "HTTP Request",
"type": "main",
"index": 0
}
]
]
},
"Register to Notion Database": {
"main": [
[
{
"node": "Freeze page_id",
"type": "main",
"index": 0
},
{
"node": "Send a text message",
"type": "main",
"index": 0
}
]
]
},
"Delay between paper summaries": {
"main": [
[
{
"node": "Return to paper loop (next paper)",
"type": "main",
"index": 0
}
]
]
},
"Prepare Notion payload (JSON)": {
"main": [
[
{
"node": "Edit Fields for Notion (incl. pdf)",
"type": "main",
"index": 0
}
]
]
},
"Analyze doc (Prompt Ultra-Pro)": {
"main": [
[
{
"node": "Merge page_id & Summary",
"type": "main",
"index": 1
}
]
]
},
"Filter recent papers (last 24h)": {
"main": [
[
{
"node": "Dedupe with Static Data",
"type": "main",
"index": 0
}
]
]
},
"Return to paper loop (next paper)": {
"main": [
[
{
"node": "Process each paper (Gemini + Notion loop)",
"type": "main",
"index": 0
}
]
]
},
"Edit Fields for Notion (incl. pdf)": {
"main": [
[
{
"node": "Register to Notion Database",
"type": "main",
"index": 0
}
]
]
},
"Code (Parse Gemini JSON in chunks)7": {
"main": [
[
{
"node": "Debug before append",
"type": "main",
"index": 0
}
]
]
},
"When clicking \u2018Execute workflow\u2019": {
"main": [
[
{
"node": "HTTP Request",
"type": "main",
"index": 0
}
]
]
},
"Append a block (adding \"block\" subtitle)": {
"main": [
[
{
"node": "Process each paper (Gemini + Notion loop)",
"type": "main",
"index": 0
}
]
]
},
"Process each paper (Gemini + Notion loop)": {
"main": [
[],
[
{
"node": "Clean page_id",
"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.
googlePalmApitelegramApi
Pro
For the full experience including quality scoring and batch install features for each workflow upgrade to Pro
About this workflow
This workflow turns the arXiv AI feed into a daily research assistant. Every morning it fetches the latest Artificial Intelligence papers from arXiv.org, deduplicates them, stores one page per paper in Notion (with metadata + PDF link), generates a Deep Research Summary of each…
Source: https://n8n.io/workflows/10519/ — 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.
AI Institutional Stock Valuation Engine with Risk Scoring & Scenario Targets
Google Sheets, XML, HTTP Request +3
Overview This is a production-grade, fully automated stock analysis system built entirely in n8n. It combines institutional-level financial analysis, dual AI model consensus, and a self-improving back
Google Sheets, XML, HTTP Request +3
A professional AI equity analysis automation built on n8n that transforms structured financial data and real-time news into disciplined, risk-adjusted price targets and actionable BUY/HOLD/SELL signal
Google Sheets, OpenAI, XML +3
QA Engineers, SDETs, and automation enthusiasts who want to stay updated with the latest testing news without manually monitoring multiple blogs and communities.
RSS Feed Read, OpenAI, Google Gemini +1
This workflow creates a daily “n8n News Radar” briefing: Pulls the latest n8n ecosystem updates from Blog, Community, GitHub Releases, and Reddit. Filters to the last 24 hours + keyword relevance. Use
Telegram, Notion, Google Gemini +2