Jev in production › Document and record classification

Ad Lab

Jev reads ad creative images directly and classifies them across seven creative dimensions in an open-source research tool (Ad Lab).

Open on GitHub ↗

The source publishes no measured number.
Use
Document and record classification
Industry
Media and social
Form
Open-source tool
Stage
In production
Listed
2026-10-26
Found via
github
Repository
artemnovitckii/ad-pattern-lab
Stars
32
Forks
5
Last push
2026-10-26
Language
JavaScript
License
MIT
Ad Lab screenshot
docs/preview.jpg in the artemnovitckii/ad-pattern-lab README, MIT; shown from GitHub.

The README opens with

Turn competitor ads into a searchable research library. Jev reads creative images directly, classifies seven dimensions, and helps you find examples worth studying before you make your next ad.

Watch the visual · Free resource + setup guide · Detailed setup · Verification · Data flow

Open http://127.0.0.1:5194. The bundled demo works without API keys. Filter format and hook, inspect individual ads, explore the findings, and replay the thumbnail wall and map together. Demo brands, labels, dates, brief scores and costs are fictional. Replay time is not measured API speed.

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for document and record classification