• English

      • Introduction

      In AI image and video generation, the right prompt is everything. For non-native English speakers, the hardest part is finding the right English phrasing. IMGPrompt removes that barrier with a native-language-first prompt editor built for Nano Banana, GPT-Image-2, Midjourney, FLUX, Seedance, Veo, Kling and other image/video models (also compatible with Stable Diffusion and DALL·E).

      With over 5,040 curated prompt tags, a clean interface, native-language translations for every term, and instant image previews for tagged entries, you can browse, click, and compose the perfect English prompt in seconds.

      Try IMGPrompt now and boost your creative workflow.

      IMGPrompt UI

      Why IMGPrompt

      • Browse in your own language, output fluent English, works with all mainstream image/video models.
      • 5,000+ curated tags across 16 main categories — styles, lighting, composition, camera, artists, and more.
      • Visual click-to-compose with hover (desktop) / long-press (touch) previews to keep you in creative flow.

      How It Works

      1. Browse 16 main categories in your native language.
      2. Drill down: main category → subcategory.
      3. Click a tag to insert it; hover (desktop) or long-press for ~0.5s (touch) to preview the example; click the thumbnail to zoom.
      4. Copy the generated English prompt into any AI tool.

      Features

      🏆 5,000+ Native-to-English Library

      16 main categories, hundreds of subcategories, and every tag paired with a native-language label so you know what you're picking.

      🖼️ Hover / Long-Press to Preview, Click to Zoom

      Entries with example images surface a preview popover (native-language name, thumbnail, description). On desktop it appears on hover; on touch devices, long-press the tag for ~0.5 seconds to open it. Click / tap the thumbnail to enter the lightbox:

      • Zoom in/out with scroll or buttons to see detail.
      • Rotate at 45° / 90° increments for multiple angles.
      • Download to save to your local reference library.
      • Keyboard: Esc closes; arrow keys navigate (where applicable).

      Closing the lightbox preserves your browsing and selection state — no flow interruption.

      On touch devices, a quick tap still adds / removes the tag — long-press is the preview gesture, so the two never collide.

      🎨 Dual Modes: Multicolor vs Monochrome

      • Multicolor uses category-based color blocks — strong visual hierarchy, great for browsing and inspiration.
      • Monochrome maximizes information density — great for quick scanning and comparison.
      • Toggle lives in the top-right of the tag section; preference is saved locally.

      🔗 Share Your Current View

      The URL carries your current browsing state as ?object=<main>&attribute=<sub>. Send the link to a teammate and they open to the exact same category — perfect for team workflows and tutorials.

      🤖 Smart Suggestions & Translation

      • Type in any of 18 languages to get up to 10 matching tag suggestions.
      • Built-in translator converts arbitrary text to English; multi-service fallback (Google → Youdao → MyMemory) with race protection.
      • 1.5s debounce to avoid spamming APIs.

      Smart Suggestion Demo

      Translation Demo

      🧩 One-Click Templates & Cleanup

      • Quick insert: portrait lighting, quality polish, negative prompts.
      • Comma-based tagging: comma input triggers auto-dedup, grouping, and chip rendering.
      • Random color swap: replace color words at random to spark fresh ideas.

      📊 Character Counter

      Real-time length counter — keep prompts under ~380 characters for better model reproducibility.

      🌓 Light / Dark Theme

      Light, Dark, and Follow-System modes, powered by the antd 6 token system — easy on the eyes at night.

      🔒 Fully Local, Fully Private

      Everything runs in the browser. No login, no upload, no backend. State persists across tab switches.

      📂 Open & Customizable Data

      Prompt data lives in open JSON (src/app/data/prompt/*.json). Edit directly or use the IMGPrompt Data Converter for bulk import.

      🖥️ Cross-Platform Client

      Prefer desktop? Grab Windows, macOS (Intel + Apple Silicon), or Linux builds from GitHub Releases.

      🌍 18 Languages

      UI and prompt database in: English, Simplified/Traditional Chinese, Japanese, Korean, Spanish, French, German, Portuguese, Russian, Arabic, Hindi, Bengali, Indonesian, Vietnamese, Thai, Turkish, Italian.