How to use Leonardo AI?
Sign up free at leonardo.ai, open Image Generation, pick a model (Phoenix or Lucid Origin for realism), type a descriptive prompt, set the number and size of images, and click Generate. Each generation costs tokens (free plan gives 150/day). Then upscale, edit, or turn favorites into video. No install needed — it runs in your browser.
Why — the first-principles explanation
Leonardo is a text-to-image tool: you describe a picture in words and a trained neural network paints it. The model learned by studying millions of image-and-caption pairs, so it maps your words onto visual patterns it has seen. That's why the prompt is your steering wheel — the more specific and visual your description, the closer the result matches your intent.
Under the hood you're choosing a model and a set of settings. Different models were trained differently: some excel at photorealism, others at anime or concept art. Settings like image size, number of images, and 'guidance' control how literally the model follows your words versus how much it improvises. Because each generation runs expensive GPU time, the platform meters it with tokens — so you learn to generate a small batch, judge it, then refine, rather than blasting hundreds of images blindly.
The workflow is iterative by design: generate, pick the best, then upscale it for detail, use in-painting to fix a hand or swap a background, or feed it into the video tool for motion. Everything happens in the browser because the heavy computing runs on Leonardo's servers, not your device — a laptop and a good prompt are all you need.
An example that makes it click
Using Leonardo is like ordering at a sandwich counter where a very fast artist makes whatever you describe. You say, 'a red fox sitting on a snowy log at sunrise, photorealistic, soft light,' and four sketches slide back in seconds. You point at the best one and say 'make it sharper' (upscale) or 'fix the tail' (edit).
The clearer your order, the better the sandwich. Say just 'a fox' and you get something generic; add the snow, the sunrise, and the camera style, and you get exactly the picture in your head. Each order costs a few tokens, like coins at the counter, so you place small orders and adjust rather than dumping your whole purse at once.
How to do it
- Go to leonardo.ai and create a free account (email or Google sign-in).
- From the dashboard, open 'Image Generation' (or 'Image Creation').
- Choose a model — Phoenix or Lucid Origin for realistic photos, others for anime or art styles.
- Write a detailed prompt: subject, setting, lighting, style, and camera details.
- Set number of images and image dimensions, then click Generate (watch the token cost on the button).
- Select your favorite, then Upscale for higher resolution or use the Editor to inpaint and refine.
- Optionally send the image to the Video tool to add motion, then download your result.
Key facts
- Leonardo AI runs entirely in a web browser — no download or install required (as of 2026-07).
- The free plan provides 150 tokens per day to generate images.
- Realistic-image models include Phoenix and the newer Lucid Origin; other models target anime, art, and design.
- Every generation shows its token cost on the Generate button before you commit.
- Beyond generation, tools include Upscale, an image Editor (inpainting), and text/image-to-video.
- Leonardo.Ai is owned by Canva (acquisition announced July 29, 2024).
An image generator popular with game artists and designers.
Affiliate link — we may earn a commission at no cost to you.▶ The 60-second explainer (script)
Want to use Leonardo AI? It's easier than it looks, and it all runs in your browser. First, go to leonardo dot ai and sign up free — you get 150 tokens a day to start. Open Image Generation. Now pick a model: Phoenix or Lucid Origin are great for realistic photos, and there are others for anime and art. Here's the key part — write a detailed prompt. Don't just say 'a fox'; say 'a red fox on a snowy log at sunrise, photorealistic, soft light.' The more specific your words, the better the picture, because the AI paints from your description. Set how many images and what size, then hit Generate — the token cost shows right on the button. Love one? Upscale it for extra detail, or use the editor to fix small things, or send it to the video tool to add motion. Generate, refine, repeat. That's the whole loop.
What authoritative sources say
People also ask
Do I need to download anything to use Leonardo AI?
No. It's a web app that runs in your browser; there's also a mobile app, but the heavy computing happens on Leonardo's servers.
Is Leonardo AI hard for beginners?
Not really. Sign up, pick a model, type what you want, and click Generate. The main skill to build is writing clear, detailed prompts.
What can I make besides images?
You can upscale images, edit them with inpainting, generate variations, and create short videos from text or from an image.
How many images can I make for free?
The free plan gives 150 tokens per day. A small image costs a few tokens, so that's typically several dozen generations daily, depending on settings.