NormalMapBench vs Leonardo.ai vs ComfyUI vs Midjourney

Pricing, plans and key features compared side by side. Updated November 2026.

NormalMapBench
🎨 Image Generation
Visit NormalMapBench →
Leonardo.ai
🎨 Image Generation
Visit Leonardo.ai →
ComfyUI
🎨 Image Generation
Visit ComfyUI →
Midjourney
🎨 Image Generation
Visit Midjourney →
Pricing model
Free
Freemium
Free
Paid
Free plan
✓ Yes
✓ Yes
✓ Yes
✕ No
Paid plans from
Completely free
$10/month (annual)
Completely free
$10/month
Who it's for
3D artists and game developers use NormalMapBench to quickly generate normal maps from heightmaps or images without installing any software. The tool runs entirely in your browser, offers real-time 3D preview, adjustable strength and scale sliders, and supports common input/output formats including PNG and EXR. Free, open source, and no account required — ideal for creating normal maps for PBR workflows, game assets, and 3D modeling projects.
Leonardo.ai fills the gap between Midjourney's artistic quality and production-ready asset workflows. The real-time generation, canvas editor, and 3D texture synthesis make it uniquely powerful for creative professionals and game developers. The free tier is incredibly generous, and even the Apprentice plan at $10/month provides serious value for anyone generating visual content regularly.
ComfyUI is the most powerful and flexible interface for AI image generation, giving users pixel-level control over every aspect of the process. While tools like Midjourney prioritize ease of use, ComfyUI is built for power users who want to build, customize, and automate complex generation workflows.
Midjourney remains the gold standard for AI image generation when artistic quality matters more than raw accuracy. Its proprietary models are tuned for cinematic lighting, atmospheric effects, and professional composition — turning even simple prompts into stunning visuals. For designers, marketers, and creators who need beautiful, commercially usable images, nothing else matches its output quality.
Key features
  • Browser-based normal map generator runs entirely in your browser with no server upload
  • Heightmap to normal map upload grayscale heightmaps or images to generate normal maps
  • Real-time 3D preview view the generated normal map applied to a 3D sphere in real time
  • Adjustable strength & scale tune the normal map intensity and detail level with sliders
  • Multiple format support input: PNG, JPG, etc.; output: PNG or EXR (high dynamic range)
  • Sobel operator uses the Sobel operator for accurate normal calculation from heightmaps
  • WebGL acceleration GPU-accelerated preview when available, falls back to CPU
  • Offline capable works without internet after initial load (service worker caching)
+2 more on the full profile →
  • Multiple Fine-Tuned AI Models Over 20 specialized models for different art styles: photorealistic, anime, illustration, 3D, concept art, oil painting, and more
  • Real-Time Generation Fast image generation with upscaling, enhancement, and variation features
  • Canvas Editor Powerful in-browser editor for compositing, masking, and refining AI-generated images
  • 3D Texture Synthesis Generate photorealistic textures for 3D models and game assets
  • Motion Turn static images into animated videos with AI-powered motion effects
  • Custom Model Training Fine-tune models on your own dataset for brand-consistent generation
  • API Access Integrate Leonardo's AI image generation into apps, workflows, and automations
  • Background Removal and Transparency Built-in tools for asset preparation and editing
  • Node-Based Workflow Every part of the generation process is a separate node; build complex custom pipelines by connecting them visually
  • Model Support Works with Stable Diffusion, Flux, and other major image generation models
  • Custom Nodes Extend functionality with community-built nodes for specialized tasks like control nets, IP adapters, and upscaling
  • Local & Cloud Run locally on your own GPU or deploy to cloud instances for more powerful generation
  • Open Source Fully open-source with an active community on GitHub and Discord
  • Batch Processing Generate multiple images with different parameters using the same workflow
  • API Access Integrate ComfyUI workflows into external applications via its REST API
  • V7 Model The current flagship model with photorealistic skin textures, correct anatomy, and a massive leap in understanding complex spatial prompts
  • Personalization Learns your aesthetic preferences over time by ranking images, creating a custom style profile so generations look unique to you
  • Character & Style Reference Upload an image of a character or art style and Midjourney maintains that exact look across new generations, ideal for storyboarding and brand consistency
  • Web Editor Browser-based interface with inpainting (changing specific parts of an image) and outpainting (expanding canvas borders) using simple brush tools
  • Niji Model A specialized sub-model (Niji 7) tuned for high-end anime, manga, and illustrative styles
  • Video Generation Generate short videos from text or images, with HD video available on Standard and above plans
  • Stealth Mode Keep images and videos private, available on Pro and Mega plans
  • Concurrent Prompts Run multiple generations at once (up to 12 fast + 3 relax on Pro/Mega)
All plans
Free $0
Full access to all features, no account or sign-up required.
Free $0/month
150 tokens/day, access to community models, basic generation features, transparent backgrounds.
Apprentice $10/month (annual)
8,500 tokens/month, private generations, premium models, faster processing, commercial license.
Artisan $24/month (annual)
25,000 tokens/month, canvas editor, 3D textures, motion, API access, priority queue.
Maestro $48/month (annual)
60,000 tokens/month, custom model training, max priority, all features, commercial license.
Open Source Free
Full ComfyUI software is free and open-source. You provide your own hardware or cloud GPU instance.
Basic $10/month
3.3 hours of fast GPU time per month (~200 images), SD video, work solo in Discord DMs
Standard $30/month
15 hours fast GPU time, unlimited Relax image generation, SD & HD video, 3 concurrent fast prompts
Pro $60/month
30 hours fast GPU time, unlimited Relax images and SD video, Stealth Mode, 12 concurrent fast prompts, 40-job repeat/permutation
Mega $120/month
60 hours fast GPU time, unlimited Relax images and SD video, Stealth Mode, 12 concurrent fast + 3 relax video prompts, built for high-volume studios
Profile views
118 on OnePlaceForAI
345 on OnePlaceForAI
378 on OnePlaceForAI
679 on OnePlaceForAI
Build a different comparison →
New AI tools, every Monday Free weekly digest — tool picks, prompts, and AI news.