Start from a format that already gets the click
FatThumb ships 16 viral-format thumbnail templates built around click patterns that have worked across millions of videos. Pick a format, add your prompt, and generate.
How it works
Three steps from zero to upload-ready.
Browse 16 viral formats
Each template is built around a proven click mechanic — reaction faces, before/after, big-reveal text, side-by-side comparison, and more. Browse by category or intent.
Customise with your prompt and Person
Select a template, write your prompt describing the specific shot, and optionally attach a Person profile. The template format guides the composition; your content drives the context.
Generate and iterate
FatThumb generates 1–4 variations fitting the template pattern with your face and content locked in. Star the best one and it is saved to your version history.
Features
Everything you need for the click.
Formats built from viral patterns
Each of the 16 templates is distilled from click mechanics observed across high-performing YouTube thumbnails — not random styles, but structured composition patterns.
Your face in the template
Combine any template with a Person profile. The template handles the composition pattern; your face reference handles the likeness. Both work together in a single generation pass.
Up to 4 template variations
Every template generation produces up to four variations. The composition pattern is consistent; the AI explores lighting, expression, background, and detail within that frame.
Reuse a template across episodes
Found a template that performs? Use it as the base for your whole series. The format stays consistent while the content and face expression change per episode.
Visual identity at scale
Running the same template pattern across a channel creates a recognisable visual series — viewers know your thumbnails before they read the title.
Every starred variation saved
Star the variations you want to keep. FatThumb's version history records every generation per prompt so you can go back to an earlier direction without regenerating from scratch.
FAQ
Questions & answers
What are the FatThumb thumbnail templates?
They are 16 pre-built composition formats based on click mechanics proven across high-performing YouTube thumbnails — patterns like reaction-face, before/after, reveal text, or side-by-side comparison. Choosing a template guides the AI's composition without limiting your prompt creativity.
Can I combine a template with my own Person profile?
Yes. Templates and Person profiles are independent inputs that work together. The template sets the composition pattern; your Person profile keeps your face consistent. Both are applied in the same generation pass.
How do I iterate once I have a template I like?
Star the variations you want to keep and they are saved to your version history. On the next generation you can stay on the same template, adjust the prompt, and compare the new batch against the starred previous version.
Can I use a template for a whole YouTube series?
Yes, and it is a recommended workflow for channel consistency. Running the same template across multiple videos creates a visual series identity — viewers recognise your thumbnail format before reading the title.
Are templates different from Inspiration references?
Yes. Templates are pre-built composition patterns included in FatThumb. Inspiration references are extracted from external YouTube thumbnails you add yourself. You can combine both — select a template and attach an Inspiration reference to blend a proven format with a specific style signal.
Can I suggest a new template format?
Template requests go to the product roadmap. The current 16 formats cover the highest-intent click mechanics. Reach out via the contact page with the pattern you want and it will be considered for a future release.
Skip the blank canvas — start from a format that works
Generate your first 5 thumbnails free — no card, no designer, your face consistent from the first run.