At DesignGen, our mission is to be the fastest path from idea to delivery, and with the Jade Lotus (v2.2) release we are taking an incremental step forward to drive this mission.
This drop centers on Create Studio and the new wave of Base, Ultra, and Ultra Max models. Everything else in this update is built to support that core experience and push you from idea to production faster and more reliably.
Jade Lotus is our reminder that the best merch workflows blend calm research, sharp design instincts, and the right automation so you can move from insight to finished product without friction.
Thanks to new scaling work on the Base inference stack, Create Studio queueing now clears roughly four times faster for every tier so prompts start streaming almost immediately.
We re-architected the Base model scheduler on top of the same infrastructure that runs Ultra and Ultra Max, which lets /v4/create sessions line up at only twenty-five percent of their previous wait time.
The new transport keeps Free, Hobbyist, Creator, and Pro accounts in sync so even mixed model runs share the same faster queue window.
Base now uses our breakthroughs in context reuse and batching to deliver steadier streaming, so you can move from idea to render without the old pause between requests.
Base is the new heart of DesignGen and powers Create Studio.
Prompt once and run Base, Ultra, and Ultra Max side by side to explore looks faster without rebuilding sessions.
This update is designed so that most of your day to day work can live inside Base, with Ultra and Ultra Max ready when you need something more specialized.
Tuned for affordability so you can create more, not less.
Generation time up to four minutes per image.
Free users can generate one hundred Base designs per month.
Base is where we expect most users to live. It is powerful enough for everyday designs, production ready concepts, and rapid experimentation, without the cost of the heaviest models.
Other models, including Premium, Ultra, and Ultra Max, are still available. For now, Base, Ultra, and Ultra Max are the main trio we see leading most work.
Smarter live trend research with Radar and Jade Lotus.
TrendTap now ships turned on by default. It runs on two new engines that replace the previous TrendTap models, which are still available inside the Legacy section for existing workflows.
Radar is the new default mode for TrendTap. It runs live searches and reasons over the results in real time to give you a focused snapshot of what matters now.
Designed for speed and responsiveness so you can quickly validate if a concept or niche still has momentum.
Ideal for daily checks, quick idea sessions, and any moment where you want a clear answer without a long research cycle.
Advanced multi-step research built on everything TrendTap has learned.
Jade Lotus performs interwoven reasoning with live data and research, using the patterns from past TrendTap usage to produce smarter, more aware results.
Better at understanding what is actually relevant and hot right now while surfacing themes, styles, and angles that make sense for real products.
Tuned to provide deeper context without the three to five minute response times of older engines, so you can plan collections or explore new niches with more confidence.
Use Jade Lotus when you are planning a collection, researching bigger bets, or want more context than a quick scan can provide.
Previous TrendTap models have been moved into a Legacy section. You can still access them if you prefer their behavior, but future improvements and tuning will focus on Radar and Jade Lotus.
Genni
Your assistant across DesignGen.
Genni is your AI assistant inside the platform. Ask Genni how to use tools, which model to choose for a job, how to set up proofs, how to clean artwork, or how to get from a rough idea to a production ready file.
You can type or speak to Genni, and use it as a guide to unlock features you may not have discovered yet. Genni is here to reduce friction and keep you moving.
Vectorize now has an Advanced Mode, giving you deeper control without changing the core product you already know.
Preview: Generate a watermarked PNG preview before committing to full vectorization, at a fraction of the full run cost, with tokens saved for one day so you can return later.
Color Controls: Auto detect colors, add/remove/edit swatches by hex, set max color count up to 256 (0 for unlimited), and adjust tolerance per color.
Shape and Output Controls: Minimum shape area, Gap Filler, SVG export draw styles, grouping options, and the ability to flatten circles/rectangles into curves for consistent behavior across tools.
Advanced Mode is especially useful for pro users preparing artwork for screen printing, vinyl cutting, laser engraving, and other production workflows.
A stronger, more capable Create Studio centered on the new Base model.
More designs for less cost, with better text and transparent output.
Faster experimentation across models and variations.
Clearer and more professional client approvals with Approve Studio.
Less manual back and forth because clients can request AI powered revisions themselves.
More realistic mockups that help designs sell online.
Smarter cleanup with MagicStick during its alpha release.
Better vector files from Vectorize with Advanced Mode.
DesignGen is moving closer to being your single pipeline from idea to design, proof, mockup, and production.
Personal Note from Alex
This release is a big one for me because it finally makes Create Studio feel like the engine I always wanted at the center of DesignGen.
The Base model, the improvements around it, and the surrounding tools like Approve Studio, Mockup Shots, MagicStick, and Advanced Mode in Vectorize are all pieces of the same goal: give you a clear path from idea to something you can actually ship.
If you use any of these features and something feels off, confusing, or almost there but not quite, I want to know. Your feedback is what has shaped this release, and it is what will shape the next one.