Leonardo AI: Features, Pricing, Models & Alternatives (2026)

Last researched: September 26, 2026. Leonardo's own pricing page returned an automated-access block during research, so figures below are cross-referenced from independent pricing trackers — verify current numbers directly at leonardo.ai/pricing.

Leonardo AI (Leonardo.Ai) is an Australian generative-AI company that builds a web and mobile platform for AI image and video generation, real-time creative canvas editing, and custom model training. Canva acquired Leonardo.Ai in July 2024; as of 2026 Leonardo continues to operate as its own product aimed at power users and professionals, while its models also power AI features inside Canva's mainstream design editor.

What Is Leonardo AI?

Leonardo AI is a generative media platform combining first-party image models (Phoenix, Lucid Origin, Lucid Realism) with access to a wide range of third-party image and video models inside one app and API. It emphasizes production-oriented workflows: a real-time "Canvas" for iterative editing, custom model/element training on a user's own reference images, upscaling, and prompt-driven variation generation, alongside a token-metered credit system that separates "unlimited" first-party usage from metered third-party model usage.

Key Capabilities

• Realtime Canvas for interactive, brush-and-prompt based image editing
• Custom model and "element" (style/subject) training from user-uploaded reference images
• Access to first-party models (Phoenix 1.0, Lucid Origin, Lucid Realism) and third-party models (FLUX.1/FLUX.2, Ideogram 3.0, GPT Image, Nano Banana, Seedream, and video models like Veo, Kling, Sora 2) through one interface
• Upscaling and image variation tools
• "Unlimited" relaxed-queue generation on first-party models for Premium/Ultimate subscribers
• Blueprint workflow execution (multi-step automated generation pipelines)
• Direct sign-in and asset hand-off with Canva accounts

Main Features

Use Cases

• Marketing and social media creative production at scale
• Iterative concept art and illustration using the Realtime Canvas
• Brand-consistent image generation via custom-trained models/elements
• Product and packaging mockups combined with Canva's design tools
• Short-form AI video generation via integrated third-party video models
• API-driven bulk content generation for apps and content platforms

Supported Models / Technologies

Leonardo's first-party line includes Phoenix 1.0 (a from-scratch foundational model, not fine-tuned from an existing architecture, reaching roughly 5-megapixel/2048×2048 output) and Lucid Origin (tuned for Full HD output, saturated color, and in-image text rendering). Beyond its own models, Leonardo's API and app expose numerous third-party image models (FLUX.1/FLUX.2, Ideogram 3.0, GPT Image, Nano Banana, Seedream) and video models (Veo, Kling, Sora 2, Seedance, Hailuo, LTX), each billed at their own token rate rather than under the "unlimited" first-party allowance. Exact maximum resolution ceilings for every third-party model routed through Leonardo are set by the upstream vendor and are not uniformly documented on Leonardo's own site.

Platforms

• Web app (app.leonardo.ai)
• iOS app
• API (Creative Engine API, pay-as-you-go)
• Canva integration (sign-in and asset hand-off)
• An Android app was not confirmed as generally available as of this writing

Integrations

Leonardo integrates directly with Canva (shared sign-in, asset flow into Canva's Magic Studio), offers an official API with Python and TypeScript SDKs, webhook callbacks, and an MCP server for agent/IDE integrations. Third-party automation platforms (e.g., Make) also list pre-built Leonardo connectors for workflow automation.

Pricing

Figures below are cross-referenced from independent pricing trackers as of September 26, 2026, because Leonardo's own pricing page blocked automated fetching during research; confirm exact current numbers at leonardo.ai/pricing.

Advantages

• Single interface/API for both first-party and a wide range of third-party image and video models
• "Unlimited" relaxed-queue generation on first-party models reduces per-image cost anxiety for Premium/Ultimate users
• Realtime Canvas offers a more interactive editing workflow than prompt-only tools
• Direct Canva account integration is convenient for users already inside the Canva ecosystem
• Paid subscribers get full commercial ownership of their generated images

Limitations

• "Unlimited" generation applies only to first-party models; premium third-party models (Veo, Kling, Sora 2, Nano Banana Pro) always consume tokens at full rate, even on the Ultimate plan
• Free-tier tokens reset daily with no rollover, and free-tier outputs carry weaker IP rights than paid outputs
• Public content (including from free users) can be reused by Leonardo/Canva under its terms, which is a consideration for sensitive creative work
• Token-based pricing across dozens of models can make true per-image cost harder to predict than flat per-image API pricing

Security & Privacy

Leonardo states that private, paid-tier content is not used to train its models without express written consent, while free-tier content may be used for training. Paid subscribers retain full ownership, copyright, and commercial usage rights over their generated images; free-tier users receive only a non-exclusive, royalty-free license while Leonardo retains underlying IP rights. Content made public (a default for free users) can be reused, reproduced, and sublicensed by Leonardo and its successors under its terms. Users can opt out of certain personal-data sale/sharing via the "Do not sell or share my personal information" mechanism or by emailing privacy@leonardo.ai. Consult Leonardo's own Privacy Policy and Terms of Service for the current, authoritative language.

Best For

Leonardo AI is best suited to teams and individual creators who want one platform and one API to reach both a strong first-party model (for cost-predictable, "unlimited" generation) and a broad menu of third-party image/video models, especially those already working inside the Canva ecosystem.

Alternatives

• Ideogram — stronger dedicated focus on in-image text rendering for design work
• Midjourney — more established stylistic/artistic community and aesthetic presets
• Stable Diffusion / Stability AI — open-weight self-hosting option outside any subscription
• Adobe Firefly — deeper integration with Adobe Creative Cloud design tools
• Recraft — comparable design/vector focus with in-image text support
• Midjourney / RunwayML — for teams that need heavier video-generation focus

Frequently Asked Questions

Who owns Leonardo AI now?

Leonardo.Ai, an Australian generative-AI company, was acquired by Canva in July 2024. It continues to operate as a standalone product for power users while its underlying models also power AI features inside Canva itself.

How does Leonardo AI's pricing work?

Leonardo uses a token-based credit system. The free plan grants 150 Fast Tokens per day (no rollover), while paid plans (Essential, Premium, Ultimate) grant larger monthly token banks that do roll over up to a multiple of the monthly grant, plus "unlimited" slower-queue generation on Leonardo's own first-party models.

Can I use Leonardo AI images commercially?

Paid subscribers retain full ownership and commercial rights to the images they generate. Free-tier users receive only a non-exclusive, royalty-free license to use their images, while Leonardo retains underlying IP rights to free-tier outputs, and any content made public can be reused by Leonardo under its terms.

What models does Leonardo AI support?

Leonardo offers first-party models including Phoenix 1.0 and Lucid Origin, plus access to third-party models such as FLUX.1/FLUX.2, Ideogram 3.0, and various video models (Veo, Kling, Sora 2) through its unified API and app, each metered separately from the "unlimited" first-party token allowance.

Does Leonardo AI have a mobile app?

Yes, Leonardo offers an iOS app in addition to its web app, letting users generate and edit images on mobile devices.

Final Summary

Leonardo AI combines its own foundational models (Phoenix, Lucid Origin) with broad access to third-party image and video models in a single token-metered platform, now operating under Canva ownership. Its tiered subscriptions trade off token allowance, concurrency, and "unlimited" first-party generation against cost, making it a flexible but somewhat complex pricing structure to fully evaluate against single-model competitors.