Blog

Thoughts on AI design tools, image to Figma conversion, vibe coding, and the future of multimodal AI design.

Research February 9, 2026 · 12 min read · BAIR, UC Berkeley

Vibe-Figma: Why Designers in 2026 Need Multimodal AI Agents to Keep Up

From CLIP and GANs to diffusion models and flow matching — a research perspective on why the history of multimodal AI converges on solving graphic design. We coin "Vibe-Figma" and "Vibe Designing" as the designer's answer to vibe coding.

Read more →
February 5, 2026 · 8 min read

AI Image Generators Are Insane in 2026 — But There's One Big Problem

AI tools now generate UIs better than most humans. But the outputs are flat images — not editable, not responsive. We built the bridge from image to editable Figma design.

Read more →
February 7, 2026 · 10 min read

The New Vibe Coding Stack: How Multimodal AI Agents Are Solving Graphic Design's Hardest Problem

Image generation, object detection, inpainting, and vibe coding — the 2026 stack for turning visual ideas into production-ready designs.

Read more →

Ready to convert images to Figma?

Try img2figma free — 4 credits, no credit card required.

Get Started Free