Features
Reverse-engineer images into prompts for Midjourney v6, DALL-E 3, Stable Diffusion, Flux.1
Scans 40+ visual parameters including lighting, depth, composition, color temperature, bokeh, atmospheric perspective
Automatically generates negative prompts for cleaner results
Model-specific output formatting with appropriate flags and syntax
Detects artistic intent and technical details like lens focal length, haze levels, color grading
Supports JPG, PNG, WebP uploads up to 10MB
Magic Enhance feature for improving prompts
Commercial license included in Professional and Ultimate plans
Credit-based system with never-expiring credits (rollover monthly)
50% off image generation credits for subscribers
One-time credit top-ups available without expiration
Batch processing up to 50 images per session (paid plans)
REST API access for automated extraction (paid plans)
Character and style extraction for concept art and branding
Text-to-image with accurate typography
Ideogram Canvas for infinite canvas editing
Fine-Tuning custom style models
Ideogram Turbo real-time generation
Ideogram API for developers
Multiple style presets (3D, painting, photo, etc.)
Image upscaling and background removal
Layerize for editing text in generated images
Batch generation and prompt templates
Search and remix community creations
Text-guided image editing (inpaint/outpaint)
Open-weight 4.0 model on GitHub (9.3B params)
Free daily credits for testing
Web-based interface (no install required)
Commercial use on paid plans
