Deconstruct
Break the reference into what creators actually need: style, composition, lighting design, color relationships, texture, lens language, mood, and transferable visual keywords.
- style
- composition
- light
- color
- lens
A visual inspiration workbench for AIGC creators
See a good image? Viko it.
When an image stops you mid-scroll, don't just screenshot it. Viko deconstructs the frame, reverses it into a usable prompt, branches new directions, and saves image plus prompt to Eagle — so inspiration leaves your bookmarks and becomes your next starting point.
fig.01 — referenceScreenshots pile up, but it's hard to say why an image works, and harder to carry its character into your own work. Viko isn't about storing images — it's about turning one image into a real creative asset.
A good reference shouldn't just be replicated. It should be understood, taken apart, translated — and kept growing.
References are hard to find again, and harder to explain. Viko breaks an image into style, composition, light and color — legible information that settles alongside its prompt.
Many tools output walls of JSON chasing 1:1 reproduction — useless for actual creation. Viko writes natural-language prompts for GPT image models, made to create with, not to copy.
An image's value lives in its style, lens feel, mood and visual order. Viko deconstructs those first, then helps you branch new directions inside the same visual system.
Viko's core revolves around three moves — understand the frame, turn it into a prompt, branch new directions.
Break the reference into what creators actually need: style, composition, lighting design, color relationships, texture, lens language, mood, and transferable visual keywords.
Output natural-language prompts and keyword phrases built for GPT image models. Never JSON. The core character stays; the words stay usable.
Generate new directions that keep the original's character: same style, different subject; same composition, different scene; same light, different story.
Everything happens on the page you're already browsing — no tab switching, no copy-paste.
A poster and a portrait shouldn't be understood the same way. Viko tailors its deconstruction to common AIGC scenarios.
Layout order, type-image relations, hierarchy and color strategy.
From understanding one image to managing an entire visual knowledge base.
Reference to prompt to new image in one line — validate prompts instantly.
Extract an image's color order and carry the same atmosphere forward.
Make complex poses and difficult perspective legible to models.
Image, prompt and deconstruction enter your local library together.
Prompts and deconstruction records stay consistent across devices.
Process several images at once — inspiration doesn't queue.