Creating a new logo shouldn't take ages.
I wanted something fast, simple and free.
So I built Logo Lattice.
100% free.
100% browser based.
100% ownership of creations.
Link in first comment π
Your message got me thinking
You came in wanting to make something simple quickly and got dropped into a complete workflow.
Changed onboarding quite a bit:
Multiple tutorials now, not just the cube one. You pick a path based on what you want to make
Added a simple circle logo tutorial that walks you through a clean logo in a couple minutes, including concentric rings, no isometric grid needed
Basically trying to get people to that first "oh nice, I made something" moment much faster
No pressure to try it again, but if you do I'd be curious whether it feels closer to what you expected. Your feedback was really useful π
Generated logos are worthless if you can't edit them
That's the whole point of Logo Lattice. AI gives you a starting point, then you move elements, tweak shapes, change colours
But our AI-generated logos used to break in frustrating ways. Layers drifted, shapes rendered invisible and elements landed wrong
Not good enough
So I rewrote the entire SVG pipeline
Now when you generate, we use @QuiverAI's Arrow model then run it through a completely rebuilt vectorisation system:
β Single coordinate system across all layers
β Correct positioning + alignment by default
β Automatic fill handling (no more invisible shapes)
β Clean editable paths instead of fragments
β SVG imports become editable layers (paths, circles, rectangles, polygons)
Every generated logo now lands as something you can genuinely work with
No broken layers. No weird positioning
Just editable logos that behave π―
You're right. "Here's a pen, draw your own damn logo" is exactly the failure mode I'm trying to avoid π
What kind of logo were you trying to make? Company type, style, rough idea?
The tutorial examples in there lean toward geometric and square. I might be overfitting to how I use the product
If there's a gap between what people expect to make in 2 minutes and what the app helps them do, I want to close it
This kind of feedback is exactly what I need to fix onboarding, so thank you for that!
Thanks for trying it and being honest
You shouldn't have ended up wondering if you needed Studio just to draw circles. Drawing, editing and SVG export are all free. Studio is just for cloud projects, AI tools and extra exports
Reading this back, I think the real issue is the app didn't get you to a quick win fast enough. If three concentric circles felt frustrating, that's a UX problem on my end
Circles are in the Shape tool (not Draw), which I clearly haven't made obvious enough
Appreciate you calling this out. Going to use it to fix onboarding π
@mrterrycarson Iβve thought about byok a little bit. Iβm not sure how I would fit it into the current monetisation model though. Iβd want to make it worth while for the user and myself, rather than just giving the service for free you know?
@_annakulina I was actually kinda impressed. I figured out I could plug my work pc (no wifi), into my personal pc (wifi) with ethernet, and then tether my home pc to my phone's 5g. It blew my mind π
Sure, there's nothing wrong with doing it that way, and it works for a lot of people. I've been doing that myself for a long time too.
But that's the workflow I'm trying to avoid. The PNG to tracing step is what takes up a lot of my energy when making logos.
Logo Lattice starts from structured vector geometry on a snapped grid, especially isometric, so edges stay parallel, exports are clean SVG, and you can actually edit the mark.
Neither way is wrong, it just depends on what suits you better π
Generated logos are worthless if you can't edit them
That's the whole point of Logo Lattice. AI gives you a starting point, then you move elements, tweak shapes, change colours
But our AI-generated logos used to break in frustrating ways. Layers drifted, shapes rendered invisible and elements landed wrong
Not good enough
So I rewrote the entire SVG pipeline
Now when you generate, we use @QuiverAI's Arrow model then run it through a completely rebuilt vectorisation system:
β Single coordinate system across all layers
β Correct positioning + alignment by default
β Automatic fill handling (no more invisible shapes)
β Clean editable paths instead of fragments
β SVG imports become editable layers (paths, circles, rectangles, polygons)
Every generated logo now lands as something you can genuinely work with
No broken layers. No weird positioning
Just editable logos that behave π―