Zagnat and Aibetpicks.ai team up to create some Mascot 3D models
Collaboration announcements in the creative tech space often feel like two ships passing in the night, each waving politely before continuing on their separate courses. This one lands differently. When Zagnat, the studio known for turning abstract brand ideas into tangible three-dimensional characters, joined forces with Aibetpicks.ai, an AI platform built around smart selection and predictive modeling, the result was a focused experiment in mascot creation that actually moved the needle on how teams approach character design for digital products.
Setting the Stage for the Partnership
Zagnat has spent the better part of a decade refining a process that starts with personality research and ends with production-ready assets that hold up under close inspection in games, apps, and marketing campaigns. Their portfolio leans toward characters that feel lived-in rather than polished to a corporate sheen. Aibetpicks.ai, on the other hand, built its reputation on sifting through large datasets to surface the most promising options quickly, whether those options are investment signals or, in this case, design directions. The overlap was smaller than one might expect at first glance, yet it proved fertile once both sides agreed to treat mascot development as a series of informed choices rather than pure artistic intuition.
The project kicked off with a shared brief that avoided the usual vague descriptors like “fun” or “memorable.” Instead, the teams defined three measurable traits the final mascots needed to exhibit: clear silhouette readability at thumbnail size, emotional range across five distinct animation states, and technical compatibility with both real-time rendering pipelines and static marketing renders. Those constraints became the filter through which every subsequent decision passed.
Breaking Down the Three Mascots
Three characters emerged from the collaboration, each carrying a distinct role within the Aibetpicks.ai ecosystem while sharing a unified visual language. The first, a compact mechanical fox named Lumen, serves as the primary interface guide. Its design balances sharp geometric planes with soft fur simulation that reacts to lighting changes. The second character, a floating orb-like entity called Prism, handles data visualization moments and needed to convey clarity without tipping into sterile minimalism. The third, a sturdy badger named Root, anchors community-facing moments and required a grounded, approachable presence that still suggested quiet competence.
The Process in Detail
Early weeks were spent mapping Aibetpicks.ai’s existing color palette and motion guidelines onto Zagnat’s character bible template. Rather than starting with mood boards filled with existing characters, the teams ran small generative experiments inside Aibetpicks.ai to surface unexpected combinations of shapes and proportions. These outputs were never used directly; they served instead as prompts that forced the Zagnat sculptors to articulate why certain forms felt off-brand. The back-and-forth produced a living document that evolved weekly, eventually locking in proportions that felt fresh yet consistent with both companies’ visual histories.
Once proportions were approved, topology work began in parallel across all three characters. Zagnat’s lead modeler emphasized edge flow that would support both high-fidelity cinematic renders and lighter real-time versions without requiring separate meshes for each use case. This decision added upfront hours but eliminated downstream headaches when marketing requested quick turnarounds on new poses. The AI side contributed a lightweight classifier that flagged potential topology issues before manual review, cutting revision cycles by roughly a third according to internal notes.
Material and Lighting Decisions
Material work revealed another area where the two teams complemented each other. Zagnat traditionally builds materials that reward close inspection, layering micro-details that only appear under specific lighting. Aibetpicks.ai pushed for materials that remained legible even when compressed for mobile thumbnails or when viewed under the variable lighting conditions of user-generated content. The compromise involved creating modular material layers: a core PBR set for hero renders and a simplified, high-contrast variant that could be swapped in automatically based on output resolution. Testing showed the simplified set retained 92 percent of the original character recognition rate at 256-pixel widths.
Lighting studies followed a similar pattern. Rather than generic studio setups, the teams captured reference footage inside actual Aibetpicks.ai workspaces, noting how overhead LEDs interacted with matte surfaces and how window light shifted throughout the day. These references informed a small library of environment maps that could be applied consistently across all three mascots, ensuring color accuracy without requiring artists to rebuild lighting for every new render request.
Verdict
The finished mascots demonstrate that structured collaboration between a traditional design studio and an AI-focused partner can produce characters that feel both hand-crafted and systematically considered. The technical constraints never overshadowed personality, yet the efficiency gains from shared tooling and rapid feedback loops are impossible to ignore. Lumen, Prism, and Root now operate as a cohesive trio that scales across contexts without losing coherence. The work sets a practical benchmark for future mascot projects that must serve both human audiences and algorithmic distribution channels at the same time.