Hi Farhan,
I hope you’re doing well.
I’m looking for a turn-key AI image generation web app, something similar in spirit to Leonardo.ai, but simpler and more focused on a few key features.
I want you to handle everything end-to-end: from frontend + backend to AI integration, subscriptions, multi-language support and deployment.
Here is what I need in detail:
1. Overall concept
I want a SaaS website where users can:
Sign up / log in
Choose a subscription plan (Free, Standard, Pro)
Generate AI images in two ways:
Text-to-Image
Image-to-Image (upload an image and transform it)
Generate multiple images at once (batch generation)
Use a prompt system similar to the “text to image” plugins on Perchance, but improved and integrated into our own site.
2. User accounts & authentication
Email + password registration and login
Password reset / forgot password
Basic account page (see plan, usage, change password, etc.)
3. Subscription plans & limits
We want at least 3 tiers:
Free plan
Very limited generations (for example a few images per day)
Optional watermark
Standard plan
Up to 16 images per batch
Monthly or yearly billing
Pro plan
Up to 60 images per batch
Monthly or yearly billing
The system should:
Store which plan the user has
Enforce limits (batch size, total generations per month, etc.)
Automatically upgrade/downgrade when payment changes
You can suggest the best way to structure limits technically.
4. Payments / billing
Integration with Stripe (or a similar provider you recommend)
Recurring subscriptions (monthly and yearly plans)
Handle:
Successful payment → activate / upgrade plan
Failed / canceled payment → downgrade or pause account
We want this fully working in production, not just a test environment.
5. AI image generation (core feature)
We need two main modes:
Text-to-Image
User enters a prompt
Option to adjust some basic settings (e.g. style, aspect ratio, number of images, etc.)
System generates images using a model like Stable Diffusion / SDXL / Flux or a similar solution via API
Image-to-Image
User uploads an image
Adds a prompt (e.g. “anime style”, “cyberpunk”, etc.)
System transforms the uploaded image based on the prompt
Batch generation
On Standard plan: up to 16 images in one request
On Pro plan: up to 60 images in one request
This should be done efficiently (queue / job system) so the server doesn’t crash
We are open to your suggestion on which AI provider / API is best (Replicate, custom SD, etc.), but it must be scalable and stable.
6. “Perchance-style” prompt system (very important)
On Perchance there is a text-to-image plugin with a nice UI where users select options and the site builds the prompt behind the scenes.
We want something similar in behavior integrated into our website:
A form / UI where the user can choose:
style
character type
mood
background
etc. (we will provide the logic / prompt templates)
The system then automatically constructs the full prompt and sends it to the AI model.
We are NOT asking you to copy Perchance’s source code, but we want you to recreate and improve that kind of plugin logic inside our app.
Improvements we’d like:
Cleaner, more modern UI
Better handling of prompt variables
Make it easy for us later to add/edit prompt templates without touching too much code
7. Better readable text in images (if possible)
If possible, we would like the system to be tuned or configured so that:
The AI can generate images where text is more readable (for example, when we want posters, covers, titles, etc.)
If this requires a specific model or approach, we’re happy to hear your suggestions.
We know this is a limitation of many models, but any improvement you can provide here is appreciated.
8. Multi-language support (10+ languages)
We want the site to support at least 10 major languages (for example: English, Spanish, Portuguese, German, French, Turkish, Arabic, Hindi, Indonesian, etc.).
Requirements:
Language switcher in the UI
All front-end text (buttons, labels, messages, etc.) should come from a translation file or i18n system
We want to be able to add/change translations later without modifying core logic
SEO-friendly URLs / meta tags per language (if feasible within scope)
You can use any standard i18n solution (for example Next.js i18n).
9. Frontend UI / UX
Clean, modern, responsive layout (desktop + mobile)
Landing page explaining the product and plans (similar idea to Leonardo.ai style: simple and conversion-focused)
Dashboard page for logged-in users:
Generate images (Text-to-Image and Image-to-Image)
View previously generated images
See remaining credits / usage
Basic account / billing page
Design doesn’t need to be super fancy like a big startup, but it should look professional and trustworthy.
10. Admin / owner controls (basic)
At minimum we need:
Ability to see users and their plans
Ability to change a user’s plan manually
See basic usage stats (how many generations, etc.)
This can be a simple admin page or a simple admin panel.
11. Deployment & “turn-key” de...