Leonardo AI

Canva's AI image and video platform: its own Phoenix and Lucid models, 20+ partner models, Blueprints and a 105MP upscaler. A free mini-course, plus how its tokens work.

What it nails

• Own Phoenix and Lucid models, plus partner models
• Free plan with daily tokens
• Pro Upscaler up to 105MP

Watch out for

• Partner models always cost tokens

Leonardo.Ai calls itself "The Creator-First Generative AI Platform". Founded in Sydney in 2022 and bought by Canva in 2024, it makes images and videos with its own models (Phoenix, Lucid Origin, Lucid Realism, Motion 2.0) and partner models like Nano Banana Pro, Seedream 4.5, GPT Image, Flux.2, Ideogram, Veo 3, Sora 2 and Kling.

Start here: the basics

1. Sign up free at leonardo.ai: you get 150 tokens a day.
2. Pick a model. Phoenix "faithfully follows your prompt, even for long, detailed instructions"; Lucid Origin is "the most adaptable and prompt-responsive model to date".
3. Build your prompt from subject ("the who or what"), context ("the where, when, or how") and style ("the overall visual aesthetic").
4. Iterate fast with Flow State: "Start with a prompt, review multiple directions, and refine until the right image emerges." It costs 1 token per image.

A photorealistic close-up portrait of a woman with red hair, in soft lighting.

Example by Leonardo.Ai · source ↗ · Leonardo's own before example, made with Lucid Origin.

A photorealistic close-up portrait of a woman with red hair, with a light dusting of freckles across her nose and cheeks, and detailed, visible skin pores, cinematic catchlight.

Example by Leonardo.Ai · source ↗ · The improved version from the same guide.

Which model to pick

Power moves

1. Phoenix for precise, detailed prompts

Leonardo's own Phoenix page shows how far a long, exact description goes: skin texture, materials, light and composition.

A minimalist studio photograph of a dark-skinned hand against a plain dark blue background. The hand is elegantly positioned, with slightly curved fingers. On the index finger is a large, smooth silver ring with an organic, fluid, pebble-like shape. The subject is wearing a light blue shirt with visible cuffs and a matte texture. The skin should appear natural: visible pores, fine lines, natural micro-creases at the knuckles, subtle variations in skin tone, and crisp, high-frequency detail. Soft, diffused studio lighting, soft shadows, high resolution, clean composition, no text.

Example by Leonardo.Ai · source ↗ · Made with Phoenix.

Ultra-realistic minimalist villa courtyard with bold terracotta walls and white cubic volumes, a narrow reflective pool, and sharp midday shadows. A lone man in a dark suit stands in shade beside a sculptural agave, framed by tree shadows and a vivid blue sky, cinematic architectural style, high detail.

Example by Leonardo.Ai · source ↗ · Made with Phoenix.

The photograph shows a close-up of an elegantly decorated tart. The base is a golden-brown, flaky crust. Topped with a generous layer of delicate, creamy yellow icing, carefully applied in swirls. The dessert is garnished with a fresh, vibrant blackberry, arranged vertically in the center. The background is a soft, neutral yellow, accentuating the tart. The overall composition is simple yet sophisticated, highlighting the textures and colors of the dessert.

Example by Leonardo.Ai · source ↗ · Made with Phoenix.

Ultra-realistic surreal landscape with a solitary tree featuring vibrant purple foliage atop a smooth white rocky hill, surrounded by scattered stones and a small reflective water pool. Clear blue sky, crisp lighting, high contrast, cinematic composition, fine detail, dreamlike yet natural aesthetic.

Example by Leonardo.Ai · source ↗ · Made with Phoenix.

A close-up, high-resolution photograph of bunches of white currants against a blue background. The composition is vertical, with the currants arranged diagonally from top left to bottom right. Each berry is translucent, revealing its seeds. Soft lighting highlights the texture and subtle color variations of the currants, from pale cream to light beige. The background is a uniform blue, creating contrast and accentuating the delicate appearance of the berries.

Example by Leonardo.Ai · source ↗ · Made with Phoenix.

2. Text in images

"Enclose the exact text you want to render in double quotation marks" and "limit your text to a few words".

A photorealistic mockup of a heather-gray t-shirt. On the chest, the text "Caffeine & Keyframes" is printed in a retro, slightly distressed, all-caps font. The text is flanked by a small, stylized coffee bean icon on the left and a keyframe diamond icon on the right.

Example by Leonardo.Ai · source ↗ · An official example made with Ideogram 3.0 on Leonardo. Keep text short: "ideally under 25 characters".

3. Lucid Origin speaks cinema

From Leonardo's help center: "Lean into lighting, lens, mood - Lucid Origin understands cinematic language", and "Describe medium (gouache, cel-shaded, vector), colour schemes, composition cues".

A lone samurai with a katana, walking through a misty bamboo forest at dusk, soft, fiery glow in the background, reflections in water, cinematic composition, highly detailed anime style, dynamic lighting, volumetric fog, intricate textures

Example by ClickUp · source ↗

4. Blueprints: one-click workflows

Blueprints are "ready-made AI workflows designed to help you achieve awesome results in just a click", such as Product Studio Photoshoot, Furnish My Space or Remove Object from Video. "Each Blueprint has its own token cost, which you can see before generating."

5. Pro Upscaler and your own models

The Pro Upscaler (May 2026) "scales AI content to true 105MP with zero distortions", with a Fix AI Artifacts switch. Elements let you "create custom elements or unique models based on your own images and style".

How Leonardo tokens work

• Unused tokens go to a rollover bank of up to 3 months, but "If you cancel your Premium plan, any unused tokens in your Rollover bank will be cancelled."
• Top-up packs "never expire".
• Unlimited "relaxed" generation runs "at a lower priority" and does not cover partner models like GPT-Image-1, Flux.1 Kontext, Ideogram or Nano Banana.

Want the partner models without a plan?

Many of Leonardo's partner models also run in AI Whisperer Lab by the run: Nano Banana Pro, Seedream 4.5, GPT Image, Flux 2, Ideogram, Veo 3.1, Kling and more.

FAQ

Sources: Leonardo.Ai · Phoenix · Pricing · Tokens FAQ · AI image prompts guide