Leonardo AI

Leonardo AI (written Leonardo.Ai by the company) is a web platform and mobile app for generating, editing and upscaling images and video. It brings several AI models together, including its own and third-party ones, and adds creative tools such as real-time generation and training your own style models. It is now part of Canva.

Quick facts

• Made by: Leonardo.Ai, a company founded in Australia in 2022 and now part of Canva
• Free plan: Yes, with 150 tokens a day. Free creations are public.
• Paid plans: Essential, Premium and Ultimate for individuals, plus team plans and a pay-as-you-go API
• Works on: Web, iOS and Android, with an API for developers
• Pricing last checked: October 2026, using several recent published sources. Token allowances and plan names change, so confirm on Leonardo’s official pricing page.

What Leonardo AI is best at

• Many models in one place: try different image models without switching between tools.
• Creative control: real-time canvas, upscaling, image editing and the ability to train personal style models.
• Game and design assets: concept art, characters, textures and product-style imagery.
• Image-to-video: turn still images into short clips on supported plans.
• Team workflows: shared token pools and private team generations.

Plans and pricing

Every image, video, upscale or edit spends tokens based on its computing cost, so the same allowance gives different numbers of results depending on the model and settings. Yearly billing is up to about 20% cheaper. One source reports the free tier is for non-commercial use, so read Leonardo’s current licence terms before using any output commercially.

Best for

• Designers, marketers and game developers who want to compare several image models.
• Creators who want to train a consistent personal or brand style.
• Teams that want a shared, managed token budget.

Less suited to

• Anyone who needs editable vector output, because it produces images, not vector files.
• Users who want one predictable model and a flat, unmetered price. Midjourney follows a different approach.
• Free-tier users who need private or clearly commercial work.

How to get started

1. Sign up for the free plan and try a few prompts across different models.
2. Check how many tokens each result costs so you can estimate a month’s use.
3. Test the upscaler, the real-time canvas and, if you want a consistent look, model training.
4. Read the licence terms before using images commercially, and move to a paid plan if you need private generations.
5. Choose a plan by comparing the token allowance against your real usage.

Prompt ideas to copy

A cosy independent bookshop at dusk, warm light in the windows, rain on the pavement, soft painterly illustration, wide shot.

Product photograph of a ceramic coffee mug on a pale oak table, soft morning light, shallow depth of field, clean minimal background.

Character concept sheet of a friendly robot gardener in three views: front, side and back. Flat colours, neutral grey background.

Isometric game icon of a wooden treasure chest, bright colours, clean outlines, transparent-looking plain background.

Limitations and things to watch

• Token costs vary by model and setting, so budgeting takes some trial and error.
• Text inside images and fine details can still come out wrong, so check the results.
• Free creations are public, and commercial use rules differ by plan.
• Plan names, allowances and which models are included change often.

Alternatives

Compare Midjourney for stylised images, Runway for video, and ChatGPT or Gemini for images created inside a general assistant.