22% of the corpus clears 2×. Start there.
99,539 videos ranked by views ÷ their own channel's median — never by raw view count.
Free to search · channel-size, duration and engagement filters on Plus
Find data-backed ideas, generate viral thumbnails, and craft winning titles.
Join 10,000+ Creators · No credit card required
Why outliers
View counts lie. A video with 500K views sounds impressive until you realise the channel averages 2 million. That's an underperformer. But 500K views on a channel averaging 50K? That's a video worth studying.
vidfora scores every video against its own channel's median, so the feed only shows ideas that genuinely overperformed — the 1-in-10 videos worth reverse-engineering.
500K views · channel averages 2M
Sounds big. Actually an underperformer.
500K views · channel averages 50K
A true breakout. Study everything about it.
outlier score = views ÷ channel median views, same format
AI Thumbnail Generator
Describe the video, get a scroll-stopping thumbnail. Boost your CTR and drive more views — no design skills needed. Every image below was generated with vidfora.








Free design studio on every account · AI generation on Plus · no credit card required
Features
Research what's already working, then package your version with thumbnails and titles built to win the click.
99,539 videos ranked by views ÷ their own channel's median — never by raw view count.
Free to search · channel-size, duration and engagement filters on Plus
Scores a channel's recent uploads against its own median, on demand.
Live fetch · no API key of your own required
Every niche with its channel count, its volume, and its biggest breakout. Ranked, not vibed.
Four of 65 shown, spread across the range — the explorer ranks them all
Proven thumbnails with the titles stripped off, so you read the packaging and nothing else.
Visual lookalikes available on a curated subset of the library
Drops into a real YouTube feed beside real outliers — home, search, sidebar and mobile.
Runs entirely in your browser — the image is never uploaded
Concepts arrive as a spec — headline, palette, layout — then you edit and export clean.
Designer free on every account · AI concepts & images on Plus (300 credits/mo)
Modeled on real outliers in your niche. The only score it returns is character count.
No invented CTR grade — YouTube truncates around 70 characters, so that's what we measure
Folder your rivals, set 2× / 5× / 10×, and see every breakout since your last visit.
5 channels on Free, unlimited on Plus · alerts surface in-app, not by email
Pricing
The outlier feed and every creator tool are free — no credit card required. Upgrade for the research suite, live analysis, and AI.
Limited-time launch offer
Extra 50% off annual — on top of the 50% you already save
Lock in the launch rate for as long as you keep your plan. When the timer ends, it's gone.
Offer ends in …
Find outliers, use every creator tool, and track your first channels.
$87/yr — save 75% vs monthly, locked in for life
$207/yr — save 75% vs monthly, locked in for life
FAQ
A video's views measured against what its own channel normally does — not against the rest of YouTube. A 7.1x means it did 7.1 times the channel's typical number, which is the signal that a topic or packaging idea is worth studying regardless of channel size.
Three choices that a naive views-over-average calculation gets wrong. The baseline is the median of the channel's other videos, so one past viral hit can't quietly raise the bar on everything that follows it. The video being scored is excluded from its own baseline, so a breakout can't dilute the very number it's being judged against. And Shorts and long-form are scored in separate pools, so a Short never looks like a 10x just because it sits on a long-form channel.
With an accumulation curve. A video three days into its life has collected roughly 28% of the views it will hold at thirty days, so scoring it against a full mature baseline would badly understate how hard it is overperforming. We scale the baseline down by that expected fraction — and scale the channel's other still-young videos up to the same thirty-day frame, so they don't drag the baseline down from the other side. A breakout shows up on day three at close to the multiplier it will settle at on day thirty. Most tools either wait a month or compare a three-day-old video against three-year-old ones and call it a flop.
No. vidfora ships with a continuously-refreshed dataset of scored videos across many channels, and the Live Channel Analyzer fetches fresh channel data on the server — no API keys or quotas to manage.
Raw views mostly measure channel size, not idea strength. A million views on a ten-million-subscriber channel is a quiet day; a million views on a fifty-thousand-subscriber channel is a breakout. Outlier scores normalize for the channel so you only see ideas that genuinely overperformed.
Yes. Cancel in one click and you keep access through the end of your billing period — monthly or annual. The Free plan never expires and needs no credit card.
Annual billing is already 50% off the monthly rate. For a limited time we take an extra 50% off that annual price — and it isn't a first-year teaser: your subscription renews at the same discounted rate for as long as you keep it. When the timer on this page runs out, new subscriptions go back to the regular annual price.
Yes. Your saved boards, tracked channels, and searches are tied to your account, visible only to you, and never shared or sold.
Stop guessing what to make next. Search the outlier database and find the ideas your audience already proved they want.
Try for FreeNo credit card required