Craftsmanship
Years in the making
Hand-drawn traits
Layers
Generation rules
Generated images
Final Vani NFTs
How Vani was Born
Vani comes to life through a hybrid generative art process, combining hand-drawn elements, algorithmic generation, iterative selection, and manual refinement. It balances algorithmic randomness with human aesthetic judgment and manual fine-tuning.
In total, this process generated approximately 3 million candidates, from which 10,000 were selected (≈0.33%), ensuring each piece is unique, consistent, and artistically refined.
Explore the Gallery for the full collection, or head to Traits to see how each trait was balanced.
STEP 01Hand-drawn Foundation
We built a master visual library of 881 unique elements across 16 primary trait types, including clothing, hairstyles, skin, expressions, hats, earrings, necklaces, head accessories, face accessories, hand accessories, body accessories, and atmospheric layers. Each element is hand-drawn in a fashion illustration style, yet built for algorithmic composability.
Inspired by high fashion, lifestyle trends, and internet memes, these elements form FVFC's "cyber-luxury" DNA. This ensures both diversity and consistency. Explore the full visual philosophy in Aesthetic.
STEP 02Algorithmic Logic
Step 02 establishes the algorithmic logic: how layers stack, rarity is managed, and how elements combine across 40 independent layers to maintain the "Future Glam" aesthetic.
Our algorithm integrates 2,000+ combination rules, each governed by 10 to 20 sub-conditions. Refined over 100+ iterative cycles and 1M+ candidate images, this logic ensures an average attribute difference of over 93%. This deep algorithmic curation guarantees absolute uniqueness across every piece in the 10,000 Vani collection.
This algorithmic curation relies on three core pillars:
Pairing rules determine which traits can coexist, ensuring every combination feels intentionally styled rather than randomly mixed.
Layer order manages 40 independent layers expanded from 16 primary trait types. This allows complex features like hair to be split, so that bangs or braids can sit in front of the body while other sections stay behind.
Rarity tuning manages 881 unique trait values, balancing common features (up to 58%) with ultra-rare traits (as low as 0.01%) to ensure both style consistency and high scarcity.
STEP 03Iterative Selection
A rigorous feedback loop over 100+ rounds refined the selection. Each round generated 10,000 candidates, with only approximately 100 (1%) manually chosen for the candidate pool.
In total, we curated the final 10,000 images from over one million (1,000,000) initial outputs, constantly refining the algorithmic rules into a consistent visual language.
Every image is filtered across our four core standards:
- Aesthetic harmony: Visual balance, color palettes, and artistic composition to ensure premium quality.
- Narrative alignment: Ensuring every character's traits align with the core Future City storyline and worldview.
- Character identity: Maintaining a diverse range of character archetypes, identities, and unique personas.
- Visual precision: Identifying and eliminating rendering errors or layer artifacts within the 40-layer stack.
STEP 04Manual Refinement
Building on the 10,000 candidates from Step 03, we manually refined specific traits like eyes, hair, and accessories to eliminate subtle aesthetic inconsistencies. This fine-tuning triggered a new algorithmic generation, expanding the pool into over two million (2,000,000) potential combinations.
From this massive secondary pool of 2,000,000 images, we conducted a final, rigorous manual selection. Only the most refined 10,000 pieces were chosen, ensuring every single Vani reaches peak visual elegance beyond standard algorithmic output.
STEP 05Final Quality Audit
We executed a final audit cross-referencing all 10,000 images against their corresponding JSON metadata. This verified the accuracy of attribute naming and ensured that every trait and rarity score is synchronized with the visual layers.
Each original artwork was exported and verified at 4,000 × 4,000 pixels, checking for resolution clarity, color accuracy, and layer stacking order to eliminate any remaining visual flaws.
Aesthetic & Collectible Value
Behind every visual lies a rigorous creative logic: 881 hand-drawn elements, 2,000+ algorithmic rules, 100 iterative cycles, and an evaluation of over 3,000,000 images. This nearly two-year project stands as a masterful blend of algorithmic precision and human aesthetic vision.
By merging algorithmic engineering with hand-crafted detail, we’ve ensured three primary collectible values: Extreme Scarcity, Aesthetic Consistency, and Artisan Precision. Our incredibly low selection rate guarantees that only the most refined pieces reach the final collection.
FVFC brings together algorithmic precision and human artistry, creating a precision-crafted visual asset built for long-term collection in the digital art market.

