Last updated: [2026-06-05]

Overview

"Image to Prompt" is a browser extension that analyzes images you explicitly select on web pages, reverse-engineers an editable text-to-image prompt, and can optionally generate new images. The developer operates no server and does not collect, receive, or store any of your data. All processing happens directly between your browser and the AI provider you choose on the extension's options page. This version contains no server-side worker or gateway.

Data stored locally

The extension uses chrome.storage.local to store the following configuration locally in your browser only:

To restore the current page's state, the extension also caches the most recent reference images, prompts, generated-image history, and analysis results in local IndexedDB. This cache stays in your browser and is never uploaded to any first-party server. None of this configuration or cache is ever written to the code repository.

Data sent to external APIs

The extension only calls an external API when you actively trigger analysis or generation:

The provider is your choice — by default Evolink (https://evolink.ai); optionally OpenAI-compatible services, Google Gemini, or a custom provider you enter. Your API key is transmitted only to that provider, as the authentication credential for your own requests, and is never sent to the developer. How each provider handles your data is governed by that provider's own terms of service and privacy policy.