AI Reverse Prompt

Right-click any image to reverse-engineer its composition and generate bilingual prompts for Stable Diffusion, Midjourney, and more.

Version 1.1.0 What’s new: - Added OpenAI, Qwen, and GLM provider integration alongside Gemini - Improved provider failure handling with clearer error reporting - Refactored background, content, side panel, and popup code into smaller modules for easier maintenance - Added saved prompt search - Added saved item tags - Added saved item categories - Added export for saved prompts in JSON and TXT formats - Fixed drawer and popup text rendering issues - Fixed Chinese text and UI string corruption in multiple areas 版本 1.1.0 新增内容: 新增了OpenA1、千问和GLM提供商集成，以及与Gemini的整合改进了提供者故障处理，提供更清晰的错误报告将背景、内容、侧边栏和弹窗代码重构为更小的模块，便于维护 新增保存的提示词搜索功能 新增保存的项目标签 新增保存的项目类别 新增保存提示的JSON和TXT格式导出功能 修复了抽屉和弹出文本的渲染问题修复了多个区域中的中文文本和U字符串损坏问题 ----- AI Reverse Prompt helps users reverse-engineer image prompts from reference images they find on the web. Main features: - Right-click any image to analyze it instantly. - Generate bilingual reverse prompts in Chinese and English. - View prompt results in a floating analysis popup near the image. - Save analyzed prompt cards in a right-side collection panel. - Reopen and copy saved prompt results later. - Support a BYOK workflow using the user's own third-party AI provider key. Current provider support: - Gemini This extension does not use a developer backend in the current version. When a user provides an API key, analysis requests are sent directly from the extension to the selected third-party AI provider.