Leonardo.Ai

Leonardo.Ai is an AI image and video generator from the Canva family for creators and teams who train their own models, edit on a canvas and build on its API.

Leonardo.Ai is an AI image generator and video generator for creators, studios and teams who need finished visuals from a prompt, a sketch or a reference picture. A concept, a product shot, a character sheet or a short clip comes out of the generator instead of waiting on an illustrator or a render. Leonardo Interactive Pty Ltd, trading as Leonardo.Ai and registered in North Sydney, makes it. The company became part of the Canva family in 2024 and continues to run the platform independently from leonardo.ai. Image Generation turns a written prompt into pictures with the Phoenix, Lucid Origin, Lucid Realism and FLUX models. Quality settings, image dimensions and a negative prompt shape the result and keep unwanted elements out. Image Guidance steers a result from an uploaded picture through Image to Image, Style Reference and Content Reference. Character Reference holds the same character across a series. The Canvas Editor masks a region and regenerates it, or extends the frame past its original edges. Realtime Canvas turns a live sketch into a render, removes backgrounds and fixes a seed so a result can be reproduced. Pro Upscaler raises the resolution of a finished image. Elements train a LoRA on a subject or a style, and personal AI models learn from a user's own dataset. The Motion models generate video from the same workspace, and a REST API exposes generation and remove.bg background removal to developers. Hobbyists who generate every day and concept artists who want a model in their own style are the core users. Game and animation studios rely on it for consistent characters and props, and marketing producers for images and short video under one subscription. Teams get shared tokens, private generations and models trained from the team's pool, and developers build on the API without opening the web app. Canva users get a generator that belongs to the same family as the design tool they already work in. One login covers the whole platform, sold as a monthly or yearly token subscription for an individual, a per-seat team plan, or the pay-as-you-go API. A typical session opens with a prompt or a reference image, passes through the Canvas Editor or Realtime Canvas for corrections, and closes in Pro Upscaler before export. The jobs it gets hired for are concept art, character sheets, product mockups, marketing visuals and social clips. Paid subscribers own the content they produce and can use it commercially. Through the API the same models run inside a pipeline, an app or a batch job that a developer controls.