HeyGen + Google Sheets: AI news videos, logged
You find a great AI article, tell yourself you’ll turn it into a quick video, and then it disappears into the “later” pile. Meanwhile, your content calendar keeps asking for consistency you honestly don’t have time to manually produce.
This hits content marketers first, but founders and newsletter operators feel it too. With HeyGen video automation, you can turn trending AI news into short avatar videos on a schedule, then have every finished link neatly logged for review.
Below, you’ll see how the workflow runs hourly, how it writes the script, how it waits for rendering, and how the final URL lands in Google Sheets so nothing gets lost.
How This Automation Works
The full n8n workflow, from trigger to final output:
n8n Workflow Template: HeyGen + Google Sheets: AI news videos, logged
flowchart LR
subgraph sg0["Schedule Flow"]
direction LR
n0@{ icon: "mdi:play-circle", form: "rounded", label: "Schedule Trigger", pos: "b", h: 48 }
n1["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/httprequest.dark.svg' width='40' height='40' /></div><br/>Dumpling AI: Search AI News"]
n2@{ icon: "mdi:swap-vertical", form: "rounded", label: "Split: Individual News Items", pos: "b", h: 48 }
n3@{ icon: "mdi:cog", form: "rounded", label: "Limit: Top 4 News Results", pos: "b", h: 48 }
n4["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/httprequest.dark.svg' width='40' height='40' /></div><br/>Dumpling AI: Scrape Article .."]
n5@{ icon: "mdi:cog", form: "rounded", label: " Combine: Merge Scraped News..", pos: "b", h: 48 }
n6@{ icon: "mdi:robot", form: "rounded", label: "GPT-4o Agent: Write Video Sc..", pos: "b", h: 48 }
n7@{ icon: "mdi:brain", form: "rounded", label: " GPT-4o Model", pos: "b", h: 48 }
n8["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/httprequest.dark.svg' width='40' height='40' /></div><br/> HeyGen: Generate Avatar Video"]
n9@{ icon: "mdi:cog", form: "rounded", label: "Wait: For HeyGen to Process", pos: "b", h: 48 }
n10["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/httprequest.dark.svg' width='40' height='40' /></div><br/>HeyGen: Check Video Status"]
n11@{ icon: "mdi:swap-horizontal", form: "rounded", label: "IF: Video Completed?", pos: "b", h: 48 }
n12@{ icon: "mdi:cog", form: "rounded", label: " Wait: Retry if Not Complete", pos: "b", h: 48 }
n13@{ icon: "mdi:database", form: "rounded", label: "Google Sheets: Log Video URL", pos: "b", h: 48 }
n7 -.-> n6
n0 --> n1
n11 --> n13
n11 --> n12
n3 --> n4
n10 --> n11
n1 --> n2
n9 --> n10
n12 --> n10
n2 --> n3
n8 --> n9
n6 --> n8
n4 --> n5
n5 --> n6
end
%% Styling
classDef trigger fill:#e8f5e9,stroke:#388e3c,stroke-width:2px
classDef ai fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
classDef aiModel fill:#e8eaf6,stroke:#3f51b5,stroke-width:2px
classDef decision fill:#fff8e1,stroke:#f9a825,stroke-width:2px
classDef database fill:#fce4ec,stroke:#c2185b,stroke-width:2px
classDef api fill:#fff3e0,stroke:#e65100,stroke-width:2px
classDef code fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px
classDef disabled stroke-dasharray: 5 5,opacity: 0.5
class n0 trigger
class n6 ai
class n7 aiModel
class n11 decision
class n13 database
class n1,n4,n8,n10 api
classDef customIcon fill:none,stroke:none
class n1,n4,n8,n10 customIcon
The Problem: Turning trending news into videos is a grind
Keeping up with AI news is already a job. Turning it into 30–60 second videos is another one. You have to find articles that are actually worth covering, open them, skim for the point, write a script that sounds human, then create the video, wait for rendering, and track the link somewhere for approval or posting. It’s not one hard task. It’s a dozen tiny tasks that break your focus and eat your afternoon. And if you skip a few days, your “video habit” basically resets.
The friction compounds. Here’s where it usually breaks down.
- You spend about 30 minutes just hunting for four decent articles, and half of them are recycled takes.
- Scripts get written in a rush, which means more edits and weaker hooks when you record (or generate) the voiceover.
- Video rendering becomes a bottleneck because you have to remember to check back and grab the finished URL.
- Links and assets end up scattered across tabs, Slack messages, and notes, so publishing turns into another round of searching.
The Solution: Hourly AI news → HeyGen videos → logged in Sheets
This workflow runs on a schedule and builds a steady pipeline of short AI news videos without you babysitting it. Every hour, it searches trending AI news (using Dumpling AI), pulls the top results, and scrapes the full article text so the script isn’t based on a headline alone. Then it merges the article content and asks GPT-4o (via a LangChain agent inside n8n) to write a casual, conversational script designed for a 30–60 second video. That script is sent to HeyGen to generate an avatar video with your chosen avatar and voice. After that, the workflow waits, checks render status, retries if needed, and only logs the final URL once the video is actually ready.
It starts with an hourly trigger, then moves through “find news” and “extract content.” Next comes script generation, followed by HeyGen video creation and a render-check loop. Finally, the completed video link is written into Google Sheets so you can review, approve, and publish on your own terms.
What You Get: Automation vs. Results
| What This Workflow Automates | Results You’ll Get |
|---|---|
|
|
Example: What This Looks Like
Say you want 4 short AI news videos per day. Manually, it’s roughly 10 minutes to find each article, 15 minutes to pull key points and write a script, and another 10 minutes to create the HeyGen video and track the URL. That’s about 35 minutes per video, or a bit over 2 hours a day. With this workflow, you spend maybe 10 minutes total reviewing the finished links in Google Sheets and deciding what to publish, while n8n handles the hourly fetching, scripting, rendering checks, and logging.
What You’ll Need
- n8n instance (try n8n Cloud free)
- Self-hosting option if you prefer (Hostinger works well)
- Dumpling AI for trending AI news search.
- HeyGen to generate avatar videos from scripts.
- OpenAI API Key (get it from your OpenAI API dashboard).
- Google Sheets to log each finished video link.
Skill level: Intermediate. You’ll be comfortable adding API credentials and editing a couple of request fields like query, avatar_id, and voice_id.
Don’t want to set this up yourself? Talk to an automation expert (free 15-minute consultation).
How It Works
An hourly schedule kicks things off. The workflow runs every hour so you’re always pulling fresh stories instead of recycling yesterday’s news.
News gets collected, capped, and expanded into individual items. Dumpling AI returns a set of trending links, n8n limits it to the top four, then processes them one-by-one so each article can be scraped cleanly.
Article text is aggregated and turned into a spoken script. The scraped content is merged and sent to a LangChain agent using GPT-4o-mini (you can switch models) to produce a casual script that fits a 30–60 second delivery.
HeyGen renders the avatar video, then n8n checks until it’s done. A Wait node pauses, a status check runs, and an If condition decides whether to log the link or wait and retry. No manual refresh loops on your part.
You can easily modify the Dumpling AI search query to target a different topic (for example “Generative AI”) based on your needs. See the full implementation guide below for customization options.
Step-by-Step Implementation Guide
Step 1: Configure the Schedule Trigger
Start by defining when the workflow should run so the pipeline automatically fetches and produces new avatar videos.
- Select Scheduled Run Initiator and set your desired schedule in Rule → Interval (e.g., hourly or daily).
- Confirm the trigger connects directly to News Search Request as the first action.
Step 2: Connect the News Data Source
Configure the news search and article scraping steps that supply content for the script.
- Open News Search Request and set URL to
https://app.dumplingai.com/api/v1/search-newswith Method set toPOST. - Set JSON Body to
{ "query": "AI Agent", "language": "en", "dateRange": "pastHour", "page": "1" }. - Credential Required: Connect your httpHeaderAuth credentials in News Search Request.
- In Expand News Items, set Field to Split Out to
newsto iterate through articles. - In Cap Top Results, set Max Items to
4to limit processing. - Open Scrape Article Text and set URL to
https://app.dumplingai.com/api/v1/scrapewith Method set toPOST. - Set JSON Body to
{ "url": "{{ $json.link }}", "cleaned": "true" }. - Credential Required: Connect your httpHeaderAuth credentials in Scrape Article Text.
query or extend dateRange to capture more articles.Step 3: Set Up Script Composition with AI
Aggregate the scraped content and generate a short, single-line script using the AI agent.
- In Aggregate Article Content, keep the Fields to Aggregate for
contentto combine article text. - Open Compose Video Script and set Text to
=Here is the topic:{{ $json.content }} Here is the news article:{{ $('News Search Request').item.json.searchParameters.q }}. - In Compose Video Script, keep the System Message that forces a single-line, conversational script output.
- Ensure GPT-4o Mini Model is connected as the language model for Compose Video Script.
- Credential Required: Connect your openAiApi credentials in GPT-4o Mini Model (credentials are added on the model node, not the agent).
Step 4: Configure Avatar Video Generation and Render Checks
Send the generated script to the avatar service, pause for rendering, and poll until the video is complete.
- In Request Avatar Video, set URL to
https://api.heygen.com/v2/video/generateand Method toPOST. - Set the JSON Body to use the script output:
"input_text": "{{ $json.output }}"and keep"width": 1280,"height": 720. - Credential Required: Connect your httpHeaderAuth credentials in Request Avatar Video.
- Keep Pause for Render connected to Check Render Status to wait before polling the render status.
- In Check Render Status, set URL to
https://api.heygen.com/v1/video_status.getand add the query parameter video_id with value={{ $json.data.video_id }}. - Credential Required: Connect your httpHeaderAuth credentials in Check Render Status.
- In Render Complete Check, set the condition to equals with Left Value
={{ $json.data.status }}and Right Valuecompleted. - In Delay and Retry, keep Amount set to
20so incomplete renders loop back to Check Render Status.
Step 5: Configure the Output Destination
Record completed video links in Google Sheets for tracking and sharing.
- Open Log Video Link and set Operation to
append. - Map the Video link column to
={{ $json.data.video_url }}. - Select the spreadsheet in Document ID and choose Sheet Name
Sheet1(gid0). - Credential Required: Connect your googleSheetsOAuth2Api credentials in Log Video Link.
Step 6: Test and Activate Your Workflow
Run a complete test to verify the data flow and video output, then enable the workflow for production runs.
- Click Execute Workflow and monitor each node from Scheduled Run Initiator through Log Video Link.
- Confirm News Search Request returns articles and that Scrape Article Text produces content.
- Verify Compose Video Script outputs a single-line script and Request Avatar Video returns a
video_id. - Check that Render Complete Check routes to Log Video Link when status is
completedand that the sheet receives the video URL. - When the test succeeds, toggle the workflow to Active to enable scheduled production runs.
Common Gotchas
- Google Sheets credentials can expire or need specific permissions. If things break, check the Google connection in n8n’s Credentials manager first.
- If you’re using Wait nodes or external rendering, processing times vary. Bump up the wait duration if downstream nodes fail on empty responses.
- Default prompts in AI nodes are generic. Add your brand voice early or you’ll be editing outputs forever.
Frequently Asked Questions
About an hour if you already have the API keys and a Google Sheet ready.
No. You’ll mostly paste API keys, connect Google Sheets, and edit a few fields like the search query and HeyGen avatar settings.
Yes. n8n has a free self-hosted option and a free trial on n8n Cloud. Cloud plans start at $20/month for higher volume. You’ll also need to factor in Dumpling AI, OpenAI, and HeyGen API usage costs.
Two options: n8n Cloud (managed, easiest setup) or self-hosting on a VPS. For self-hosting, Hostinger VPS is affordable and handles n8n well. Self-hosting gives you unlimited executions but requires basic server management.
Yes, and it’s straightforward. Change the query inside the “News Search Request” (Dumpling AI search) to topics like “Generative AI” or “Autonomous Agents,” then update avatar_id and voice_id in the “Request Avatar Video” node to match your brand. You can also swap GPT-4o-mini for a different OpenAI model in the chat model node if you want a different tone. Many teams also adjust the schedule so it runs daily instead of hourly once they find a cadence they like.
Usually it’s an expired or incorrect API key in the HTTP Header credential used by the HeyGen request. Double-check the credential attached to the “Request Avatar Video” and “Check Render Status” HTTP nodes, then confirm your HeyGen account has access to the avatar_id and voice_id you’re sending. If it fails only sometimes, you may be hitting rate limits or polling too quickly; increasing the wait time before the status check often stabilizes it.
It depends on how often you run it and your plan limits. On n8n Cloud, you’re capped by monthly executions, so hourly runs plus retries add up. If you self-host, there’s no execution cap, but you’re still limited by your server and by HeyGen/OpenAI/Dumpling AI throughput. Practically, most teams start with hourly or every few hours, then tune the schedule once they see how many videos they actually publish.
For this workflow, n8n is usually a better fit because it handles branching (the render-complete checks and retries) without turning your bill into a surprise, and it’s comfortable with multi-step HTTP request chains. Zapier or Make can still work, but you’ll often end up juggling extra steps for scraping, merging content, and polling HeyGen’s render status. Also, self-hosting is a big deal if you want high volume without paying per task. If you’re trying to keep it to a simple “one link in, one link out” flow, Zapier may feel simpler. Talk to an automation expert if you want help choosing.
Once this is running, you stop “trying to stay consistent” and you just are. The workflow handles the repetitive parts so you can focus on what to say, what to publish, and what to skip.
Need Help Setting This Up?
Our automation experts can build and customize this workflow for your specific needs. Free 15-minute consultation—no commitment required.
