Leonardo AI

About

Leonardo AI carved out its niche by targeting game developers, digital artists, and creative professionals who need more than just pretty pictures. Now owned by Canva (acquired in 2024), it offers a mix of proprietary models (Phoenix 1.0, Kino XL) and third-party models (Veo 3, Sora 2, Kling, Seedance) in a single platform — giving you access to the best of multiple AI engines without juggling subscriptions. The AI Canvas is the standout feature. It's an advanced editing workspace that supports inpainting, outpainting, image expansion, and detailed region adjustments. You can select any area of an image and regenerate just that section while preserving everything around it. Combined with real-time generation that updates as you sketch, it creates a workflow closer to digital painting than prompt engineering. 3D texture generation is where Leonardo has no real competition. It generates UV-mapped albedo, normal, and roughness maps from text prompts or reference images. For game developers and 3D artists, this eliminates hours of manual texture work. No other consumer AI image platform offers production-ready 3D textures. The free plan gives you 150 daily tokens that reset every 24 hours — enough to generate 30-50 images depending on settings. The Apprentice plan at $10/month provides a larger token pool. Artisan at $24/month and Maestro at $48/month add more tokens and priority access. On Premium and Ultimate plans, when your token pool hits zero, you switch to Relaxed Generation instead of being locked out — effectively unlimited if you can wait. Video generation is included through partnerships with Veo 3, Sora 2, and other models. You can generate short videos from text or animate static images directly in the platform, though video features consume tokens faster than image generation. The affiliate program through Impact.com offers 60% of the first month's payment for every referred paid subscriber, with a 30-day cookie window. That's one of the most generous commissions in the AI tools space. The main weakness is consistency. With multiple models available, output quality varies depending on which model and settings you choose. New users can spend time figuring out which model works best for their use case. The interface also has a steeper learning curve than simpler tools like DALL-E 3 or Ideogram.

Key Features

• AI Canvas with inpainting, outpainting, and region-specific regeneration
• 3D texture generation — UV-mapped albedo, normal, and roughness maps
• Multiple AI models: Phoenix 1.0, Kino XL, Veo 3, Sora 2, Kling
• Real-time generation that updates as you sketch on canvas
• Video generation from text prompts and image animation
• Relaxed Generation for unlimited output when tokens run out
• LoRA model fine-tuning for custom styles and characters
• API access starting at $9/month for developer integrations

Use Cases

• 1Game developers generating 3D textures and concept art for assets
• 2Digital artists using AI Canvas for iterative editing and composition refinement
• 3Creative studios accessing multiple AI models in a single platform
• 4Content creators producing images and short videos from the same tool
• 5Developers integrating AI image generation via Leonardo's API