Promptoscope — Image to Prompt JSON

Analyze an image visible in Chrome and generate a structured visual recipe with a reusable prompt JSON.

Promptoscope turns a public web image into a structured visual recipe and a reusable prompt JSON. Right-click an image in Chrome and let Promptoscope analyze its visible visual characteristics: subject, setting, composition, lighting, colors, style, camera feel, details, and visible text. The result is presented in a side panel and can be copied as: • A reusable image-generation prompt • A structured JSON description • A visual breakdown of the main elements detected in the image Promptoscope is designed for designers, creative professionals, marketers, researchers, and anyone who wants to understand or recreate a visual direction more quickly. HOW IT WORKS 1. Right-click a public image on a web page. 2. Select “Analyze with Promptoscope”. 3. Review the image analysis in the Chrome side panel. 4. Copy the generated prompt or JSON structure. MODEL USED Promptoscope uses Gemini 3.1 Flash-Lite, a multimodal Google model capable of interpreting images and generating structured text. The model receives the selected image together with a dedicated visual-analysis instruction. The result is validated by the analysis server before it is displayed in the extension. Promptoscope does not retrieve the original prompt used to create an image. It generates a plausible interpretation based on the visible characteristics of the image. PRIVACY AND DATA HANDLING To process an analysis, the selected image URL and image content are temporarily sent to the dedicated analysis service, which uses Gemini 3.1 Flash-Lite. Promptoscope does not permanently store the analyzed image. Technical information such as the installation identifier, usage counters, processing time, request status, and IP address may be used to operate the service and prevent abuse. Recent results may be stored locally in the extension to avoid repeating an identical analysis. They are not stored in a server-side history. Do not use Promptoscope with confidential, private, restricted, or sensitive images unless you have verified that you are allowed to process them. LIMITATIONS The beta version is free within the following limits: • 5 analyses per day and installation • 20 analyses per day and IP address • One analysis at a time • At least 10 seconds between analyses • A global service limit of 500 analyses per day The extension currently supports: • Public HTML images • JPG, PNG, and WebP formats • Images up to 5 MB • Images up to 1,200 pixels on the longest side The following are not supported in this version: • CSS background images • Screenshots or page areas • Local files • Images behind a login • Videos • Animated GIFs • PDFs • Multiple-image analysis The model may misunderstand an image, misread text, or infer details incorrectly. Always review the generated result before using it in a professional, commercial, or published project. PERMISSIONS Promptoscope requests only the permissions required for its core features: • Context menu: add “Analyze with Promptoscope” when right-clicking an image • Side panel: display the analysis next to the page you are viewing • Storage: keep consent, quota state, and recent results locally • Clipboard: copy the generated prompt and JSON • Limited network access: send the selected image to the dedicated analysis service No account is required.