AI-powered attention prediction for Figma. 9 perceptual metrics grounded in vision neuroscience. Know your design's impact in seconds.
No data needed. No live users required. Predict attention before a single pixel goes live.
Pick any frame in your Figma file. Desktop, mobile, poster — any design type.
Our AI model (trained on 1,980 real eye-tracking sessions) predicts where humans will look.
Heatmap overlay, 9 perceptual scores, actionable suggestions, and attention timeline.
Each metric is grounded in published neuroscience research — not arbitrary heuristics.
Visual hierarchy strength. Zipf's law + Miller's 7±2 cognitive load.
Reading pattern compliance. F-pattern, Z-pattern, Gutenberg.
Foveal efficiency. The eye's 2° high-acuity zone.
Optical center alignment. 3.5% above geometric middle.
Proximity, alignment, figure-ground separation.
Retinal DoG, Weber contrast, CSF frequency analysis.
Eye movement cost. Scanpath + Inhibition of Return.
Helmlab MetricSpace. 18% more accurate than CIEDE2000.
Eccentricity-dependent acuity outside the fovea.
Beyond heatmaps — a complete attention analysis toolkit.
JET colormap overlay showing predicted visual attention. Toggle heatmap and focus map.
1s → 3s → 7s attention evolution. Three separately trained models.
Select two frames, compare scores side by side.
Check how much attention each UI element captures.
Actionable suggestions: hierarchy, CTA visibility, layout balance.
Original dimmed except high-attention regions.
Not another rule-based heuristic tool.
| Heatmap | Attention Insight | Rule-based | |
|---|---|---|---|
| AI saliency prediction | ✓ | ✓ | ✕ |
| Fine-tuned on UI data | ✓ | ✕ | ✕ |
| 9 neuroscience metrics | ✓ | ✕ | ✕ |
| Attention timeline | ✓ | ✕ | ✕ |
| Element attention | ✓ | ✕ | ✕ |
| Helmlab color | ✓ | ✕ | ✕ |
| A/B comparison | ✓ | ✓ | ✕ |
| Free tier | ✓ | ✕ | ✓ |
Free to use. No account needed. Just install and analyze.
Install Figma Plugin — Free