Colection for agate work.
AI & ML interests
Logo and brand identity generation, image captioning, preference and reward modeling for design
Recent Activity
Organization Card
LogoLabs designs custom logos and brand identities from data. We study how brands are drawn, what their briefs ask for and which choices hold up, and turn a short conversational brief into original concepts, refined with you: a logo and full brand guidelines in days, not weeks.
The research behind that work is open. These are the models, tools and data we build it on, and they run on your own machine or in your browser.
Vectoriser · Apache-2.0
Inkvec
Turns a PNG, JPEG or WebP logo into a clean SVG: edges to a fraction of a pixel, the fewest curves that match,
gradients and transparency kept. Try Studio Lite in the browser.
Text to image · MIT
Agate Preview 001
A 260M-parameter model trained from scratch in 145 GPU-hours. GenEval 0.550, level with SDXL,
and it runs in your browser on WebGPU.
Dataset · gated
LogoBrief-10K
10,000 brand logos with their original SVGs, style tags, motif descriptions and full design briefs,
sampled by web popularity and checked for AI-training opt-outs.
Also: Agate WebGPU · Agate for ComfyUI · inkvec-denoiser-001 (JPEG and AI-damage cleanup) · inkvec-sr-001 (4× upscaling for logos) · Inkvec desktop app
logolabs.org · Contact · GitHub · Inkvec docs