I left Figma to build a diffusion-based UI design tool
Details
- External ID
- 48995754
- Source
- HN
- Company
- —
- Product
- I left Figma to build a diffusion-based UI design tool
- Website domain
- diffui.ai
- Launched
- July 21, 2026
- Cohort
- —
- Upvotes
- 80
- Upvotes percentile
- 0.8936678614097969
- Tags
- —
- Fetched at
- Sept. 7, 2026, 9:26 p.m.
- Updated at
- Sept. 7, 2026, 9:26 p.m.
Description
Hey hn, this has been something I've been working on for the last few months and is finally robust enough to really show off.I've been pretty tired with the design outputs of LLMs for a while, and I've always thought diffusion offered much more creative / on brand design outputs, even before they were able to render text.I had enough conviction for this to leave my role over at Figma to build Diffui. The goal is to allow for you to design your full web app as quickly as possible, in a figma-like interface, and then hand that off to an agent to build. The page shows some interactive examples. Happy to be an open book here and chat about the approach, the process, etc!
Enrichment
- Theme
- indie hacker passion projects
- Vertical
- Horizontal
- Function
- Content generation
- Audience
- B2C
- AI stance
- AI feature
- Project type
- Commercial product
- Normalized one-liner
- diffusion-based ui design tool
- Manually corrected
- False
Could you build this?
No Building a diffusion-based UI design engine requires fine-tuning or training latent diffusion models specifically to generate structured, vector-ready UI components and layouts rather than blurry flat images.
What it would actually take: The system requires a custom diffusion pipeline (likely fine-tuned on vast datasets of UI wireframes, design tokens, and components), paired with vectorization or layout extraction models that convert latents into layered Figma-like canvas elements. It demands a dedicated GPU training infrastructure, millions of curated UI design files, and deep research expertise in generative vision models conditioned on structured design graphs.
Discussion
13 comments analyzed.
Competitors mentioned: Claude Design, Claude (for design/implementation tasks), Traditional designer + developer workflow
Concerns raised: Token budget efficiency compared to alternatives, Whether diffusion model output quality justifies workflow changes
Feature requests: MCP integration improvements for seamless agent handoff, Export/version control for brand profiles across clients, Built-in normal/depth map generation UI (currently requires manual Claude prompt)
Competitors
Other products that read as similar to this one — 150 launches clear the similarity bar, closest 8 shown.
Attention rank: #18 of 151 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 265 days after the earliest competitor.
- Mowgli · hn · 2026-02-28 · 6 upvotes · similarity 0.50
- I spent 4 years building a UI design tool with only the features I use · hn · 2026-02-06 · 386 upvotes · similarity 0.48
- CoFoundMe · hn · 2026-08-18 · 5 upvotes · similarity 0.46
- Diffusion PDF · hn · 2026-08-13 · 5 upvotes · similarity 0.45
- Foundry: a Markdown-first CMS written in Go · hn · 2026-03-27 · 31 upvotes · similarity 0.43
- I made a browser TUI with its own layout engine · hn · 2026-07-07 · 6 upvotes · similarity 0.43
- Sunwet · hn · 2026-03-20 · 5 upvotes · similarity 0.43
- I built a browser-based 3D editor since I didn't want to learn Blender · hn · 2026-03-01 · 8 upvotes · similarity 0.43
Other launches for this product
- No other launches for this product.
Same idea, different domain
Nobody's really built a content generation tool for Government yet.