VidiScoopMulti-source discovery
Google News search feeds scoped to exactly the last 24 hours, Google News topic sections, curated outlet feeds chosen for your sections and region, regional Google Trends, and any RSS feed you paste in yourself.
Home Features
Everything between “something happened” and “upload”
VidiScoop is not a trend list and not a prompt box. It is a research engine that understands which channel you are planning for, and a packaging engine that writes for that channel specifically.
Discovery
Find the story, not the noise
Most tools show you what is trending on the internet. VidiScoop answers a narrower and far more useful question: what happened in the last day that is worth a video on this channel.
Duplicate collapsing
Twelve outlets reporting the same event become one story with a coverage count of twelve — which is itself one of the strongest signals that a story is real and breaking wide.
Rarity-weighted relevance
A keyword like “movie” matches everything and is worth almost nothing. A keyword like “x-men” matches the thing you actually cover. Relevance weights each keyword by how rare it is across the scan.
Exponential freshness decay
Freshness does not fall off a cliff at a fixed hour — it decays continuously, so a 40-minute-old story is meaningfully ahead of a four-hour-old one instead of both being “today”.
Trend confirmation
Regional Google Trends is checked as a separate signal, so a story that is quietly reported but loudly searched still surfaces.
Source authority
A wire service carries more weight than an aggregator republishing it. Authority is a small but decisive tiebreaker between two otherwise identical stories.
Packaging & workflow
Turn the decision into an upload
Choosing the story is half the job. The other half is the title, the description, the tags and the thumbnail — and that is where most channels quietly lose their views.
Full video packages
Five titles, a description, tags, hashtags and three thumbnail briefs — written from the story and your channel’s positioning, not from a generic template.
Bulk generation
Select ten stories, generate ten packages in one action. Anything already packaged is reused instead of regenerated, so you never spend credits twice on the same story.
Follow-up detection
A story you already covered that resurfaces with nothing new goes to “Already covered”. If it genuinely develops, it comes back to the main list flagged as an update.
YouTube competition check
Optionally check how many videos already exist for a story before you commit, so you can tell an open lane from a crowded one.
Per-channel workspaces
Niche, keywords, language, region, sections, custom feeds and minimum score are all set per channel. Switching channels switches everything.
CSV and Markdown export
Hand the day’s packages to a writer, an editor or a production sheet without anyone needing an account.
Bring your own engine
Four AI providers, or none
The built-in engine runs the whole product with no API key at all. Add a key and the model re-reads every story in your channel’s context to sharpen the ranking, and writes the packaging properly instead of filling a template.
- Google Gemini — has a free tier, the easiest place to start.
- Claude — strongest at nuanced editorial judgement.
- OpenAI — familiar, widely available.
- OpenRouter — one key, many models, including free ones.
Keys live server-side and never reach the browser. If a provider call fails, the built-in engine takes over and the scan still completes.
What changes when you add a key
| Capability | Built-in | With a key |
|---|---|---|
| Scanning & clustering | Full | Full |
| Five-signal scoring | Full | Full |
| Editorial re-ranking | — | Yes |
| Titles & descriptions | Templated | Written |
| Thumbnail briefs | Templated | Written |
Feature questions
The detail behind the list
How does story clustering work?
Every article pulled from your feeds is normalised and compared against the others by headline and entity overlap. Reports of the same event collapse into one story, and the outlets that ran it become that story’s coverage count. This is what turns a feed of 300 articles into a list of roughly 20 distinct things that happened.
Can I trust the score, or should I read every story anyway?
Treat it as a ranking, not a verdict. The score exists to put the five or six stories worth your attention at the top of the list instead of scattered through 300 articles. Every story shows its individual signal bars, so you can see whether a high score came from real multi-outlet coverage or just a keyword match, and override it in a second.
What exactly is in a generated package?
Five title options written to different angles, a full YouTube description with the context a viewer needs, a tag list, hashtags, and three thumbnail briefs describing the shot, the text overlay and the emotional beat. Packages are saved permanently against the story and can be exported as CSV or Markdown.
Does the YouTube competition check cost extra?
It uses your own YouTube Data API key, which is free up to 10,000 units a day from Google. A scan checks the top stories only — roughly 1,500 units — so the free quota comfortably covers several scans per day.
Your next story is already breaking
See it on your own channel.
Add one channel, run one scan, and judge the ranking against what you already know about your niche.
No credit card required · Your first ranked scan takes about a minute