Free tool

The prompt generator for your AI girlfriend

Build the perfect character for your AI companion, optimized for your target app. No more failed prompts.

  • ✅ 100% free
  • 🔒 No signup
  • 🇫🇷→🇬🇧 Clean English prompt
  • 🔞 18+

To generate an image — or a video, which always starts from a seed image — you need a prompt. The catch: a good prompt is long to write, and it's more reliable in English (the AI's language). Here you describe in plain words, the tool hands you the clean English prompt, ready to paste.

Loading the tool…

Why a good prompt changes everything

On an AI companion app, the character you get is only worth what you put in up front. The engine doesn't "guess" who you want — it plays what you describe. A vague prompt gives a generic character that answers off-topic, forgets its own personality and breaks immersion within three messages. A precise prompt — coherent appearance, personality in keywords, defined tone of voice, clear opening context — gives a character that holds its role over time. The difference between a failed creation and one that hooks you is almost entirely decided at this step.

The image prompt: describe, don't "pick"

On most apps, creating the right looks means digging through the explore tab, choosing a preset model, then fumbling with image generation. It's slow and frustrating. The direct approach: describe precisely what you want — render (realistic or anime), origin, body, hair, eyes, skin, outfit, pose, framing, setting — and paste a clean prompt into the generator. Our tool assembles all of it into one optimized line, in the right order, with the right tags. You fill it once, you get an image prompt ready to paste — and you tune the intensity by one notch.

A video always starts from an image

An AI animation or video doesn't come from nowhere: it starts from a seed image. So before the video, there's always an image prompt to nail — and that's where people get stuck. The prompt is tedious to write, and it's more reliable in English, the language these AIs are trained on. When you're not a native speaker, you translate badly, you improvise, and the result disappoints. That's exactly what the tool fixes: you fill it in your own language, it assembles a clean, ordered English prompt — your image base for a photo, or the starting point for a video.

The 3 mistakes that ruin a creation

1. Vagueness. "A pretty, nice girl" tells the engine nothing. Give concrete details: hair color and cut, type of gaze, clothing style, 3-4 sharp character traits.

2. Wrong format. Pasting a narrative block into a "Personality" field built for keywords (or vice versa) dilutes the signal. Respect the app's format — that's exactly what the tool handles for you.

3. A sloppy opening. The first message sets the tone for the whole relationship. Dialogue alone falls flat; a short action narration followed by a line sets the scene and pulls you in far better.

What is a negative prompt?

The main prompt describes what you want. The negative prompt lists what you do NOT want: deformed hands, blur, watermark, text, anything that degrades the result. You paste it into the app's « Negative prompt » field, next to the main prompt. It's often what turns a decent image into a clean one, and the tool gives it to you ready to paste under your result.

Which prompt for which app?

The right prompt changes per app. For images, the tool outputs an optimized visual prompt to paste into OurDream, Xotic AI or Promptchan — and it works just as well in ComfyUI, Automatic1111 or any Stable Diffusion generator. You generate here, you paste into the app, it's already in the right format.

100% free tool, no account, nothing sent to a server: what you type stays on your device. Adults only (18+). Some terms (any reference to a minor, etc.) are blocked and prohibited. You alone are responsible for how you use your prompts: follow your country's laws and the terms of the app you use.