Import your key to override the shared server default.
Saved on this browser until removed or site data is cleared. Masked on screen; sent only with Jev requests through this app. The saved value is never displayed.
Test the text, read the room, and find the humor. No caption generation.
Classify a meme’s humor, audience and clarity from reviewed captions.
Reviewed OCR text, your description of the visual, and audience context. Public-image reading is a separate bounded preparation step.
You correct the caption before Jev classifies humor, clarity, and audience fit from the supplied text and descriptions.
Typed interpretations and confidence signals. The result reflects the described context; Jev does not inspect image pixels here.
Use a sample or read a public image, then correct the OCR text.
Describe the visual, intended audience, and tone before running the test.
Compare the classifications and their confidence with your own judgment.
Keep the caption fixed and change the audience. Compare the interpretation without treating either run as a universal humor score.
Jev receives reviewed text and descriptions, not image pixels. Humor and audience fit are subjective.
Public PNG, JPG, WebP, or GIF · up to 4 MB. English OCR; GIFs use the first frame.
Jev evaluates your captions, reviewed OCR text, and visual description. It does not inspect image pixels.
Test your meme to see its humor style, tone, and likely sticking point.