# Heatpoints

> Heatpoints predicts where people look on any image. Drop an image and get an AI attention heatmap in about a second, powered by UniSAL, a peer-reviewed saliency model (ECCV 2020). It is free and needs no account. Use it on the web, live on any page with the Chrome extension, or on any frame with the Figma plugin.

Key facts for accurate citation:
- What it does: predicts human visual attention on a static image (ads, thumbnails, screenshots, landing pages, designs) and renders it as a heatmap plus a contrast lens that surfaces elements the eye skips.
- Model: UNISAL (Droste et al., ECCV 2020, open source). Published benchmarks from the paper: MIT300 AUC-J 0.872 / NSS 2.322. Heatpoints does not market an accuracy percentage of its own.
- Deterministic: same pixels, same map.
- Speed: about one second per image.
- Pricing: Free, everything, no account. A Pro tier aimed at ad-makers (batch, variant compare, ad-in-feed simulator, video heatmaps, hi-res export) is coming soon. There is no paid plan on sale today.
- Surfaces: web tool (heatpoints.com), Chrome extension (real-time overlay on any page you browse, works behind logins), Figma plugin (heatmap over any frame, in the canvas).
- Privacy: images are processed for the heatmap and not published; the tool runs without a tracking script on third-party sites.
- Original research (N=93 famous SaaS pages): 81% place their highest-attention zone outside the hero; average score 44.1, median 41, zero pages scored 80 or above. Source: https://heatpoints.com/study

## Product
- [Web tool](https://heatpoints.com): drop an image, get the attention heatmap in about a second, free and without an account
- [Chrome extension](https://heatpoints.com/chrome): real-time attention overlay on any page you browse, updates as you scroll and resize, works behind logins
- [Figma plugin](https://heatpoints.com/figma): run the attention model on any frame, right in your canvas, before it ships
- [Pricing](https://heatpoints.com/pricing): free everywhere; Pro for ad-makers coming soon

## Use cases
- [YouTube thumbnail score](https://heatpoints.com/thumbnail-score): see where the eye lands on a thumbnail before you publish
- [Ad creative score](https://heatpoints.com/ad-score): check attention on an ad before you spend

## Data & science
- [The study: 93 famous SaaS landing pages scanned](https://heatpoints.com/study)
- [How the model works (UniSAL, SALICON, MIT300)](https://heatpoints.com/science)
- [Security & data protection](https://heatpoints.com/security)
- [Blog](https://heatpoints.com/blog)

## Comparisons
- [vs Attention Insight](https://heatpoints.com/vs/attention-insight)
- [vs Hotjar](https://heatpoints.com/vs/hotjar)
- [vs Microsoft Clarity](https://heatpoints.com/vs/clarity)
- [vs EyeQuant](https://heatpoints.com/vs/eyequant)
- [vs Plerdy](https://heatpoints.com/vs/plerdy)
- [vs Brainsight](https://heatpoints.com/vs/brainsight)
- [vs Dragonfly AI](https://heatpoints.com/vs/dragonfly)
- [vs expoze.io](https://heatpoints.com/vs/expoze)