AI drawing games start with nothing more than giving a few words or a photo to the free version of ChatGPT, the Gemini app, Microsoft Copilot, or a similar tool. For each of the 10 ideas below, this guide lists which free tool to use, the steps, and what to type. It also covers the rules for games that use photos and the age requirements for playing with kids, checked against each company's terms and official help pages (checked on November 14, 2026).
Key points
- Tools
- The free versions of ChatGPT, the Gemini app, and Copilot can make images. Canva is good for finishing touches, and Google's AutoDraw is free for doodles. None of these companies publishes a specific number for free use.
- Games with photos
- Only use photos of yourself, your pets, people who said yes, and your own room. Copilot's terms forbid creating or editing images of other people without their permission.
- Playing with kids
- For children under 13, an adult operates the tool from the adult's own account. Transforming photos (image editing) in the Gemini app requires being 18 or older.
Each company's terms and official help pages checked on November 14, 2026
Games with words only
③ character trip, ⑤ stickers, ⑥ four-panel comic, ⑦ "what if" mashups, ⑧ coloring pages, ⑩ drawing duel
- From 13, teens can also play with their own account on services whose terms allow it
- The thing to watch: do not ask for real people or existing characters
Games with photos or doodles
① doodle makeover, ② new style for a photo, ④ a child's drawing as a storybook page, ⑨ room makeover
- Transforming photos in the Gemini app requires being 18 or older
- First check the photo for faces or anything that reveals an address
Contents
This is a guide to playing, not a tool comparison or an in-depth course. To pick a tool by image quality and price, see the image generation AI tools comparison. For prompt patterns and rights explained step by step, see Getting started with AI image generation.
1. Free tools and their conditions
Every game in this guide can be played with the free versions of the six tools below. ChatGPT, the Gemini app, and Copilot make pictures and refine them through conversation. Canva turns finished pictures into cards and posts. Adobe Firefly lets you try models from several companies. AutoDraw turns lines you draw by hand into illustrations.
| Tool | What you can do for free | Free limit | Age requirement |
|---|---|---|---|
| ChatGPT | Image creation is available on all plans. You can also attach a photo and ask to transform it. | No number given for the limit. | 13 or older. Under 18 needs permission from a parent or legal guardian. |
| Gemini app | Creates images with Nano Banana 2. You can also start from a template. | A daily limit (no number given). Free downloads are 1K. | 13 or older to create images (varies by country). 18 or older to edit images. |
| Microsoft Copilot | Creates images from text, and you can upload an image and ask it to change it. | The official help and the terms give no number. | 13 or older (older in some countries, depending on local law). |
| Canva | "Create an image with Canva AI" with a limited allowance. Add text to a picture and finish it as a card or post. | A monthly allowance. Pro gets 10 times the free amount. Once it runs out, short pauses appear between generations. | Under 13 only through Canva Education for schools. |
| Adobe Firefly | Firefly Free gives limited access to models from Adobe, Google, OpenAI, and others. | Free daily generations (no number given). | 13 or older for an Adobe ID. Minors need consent from a parent or guardian. 18 or older to use non-Adobe models. |
| AutoDraw | Guesses what you are trying to draw from your lines and offers drawings by illustrators to replace them. | Works in the browser and is free (Google's introduction). | Not stated on the introduction page. Children should use it with an adult. |
Sources: Images in ChatGPT, OpenAI Terms of Use, Generate & edit images with Gemini Apps, Gemini Apps limits, Using image generation in Microsoft Copilot, Copilot Terms of Use (effective August 18, 2026), Canva AI access, Canva Terms of Use, Adobe Firefly plans, Adobe General Terms of Use, Adobe Generative AI Product Specific Terms (effective April 23, 2026, section 7.2), Google's introduction to AutoDraw.
None of these companies gives a number for how many images you can make for free. The Gemini app's help page says that generating media such as images uses more of your allowance than simple chats. If one game means remaking many pictures, spreading the work across tools helps you avoid hitting a limit. Tools with no free version, such as Midjourney, are not used in this guide. Paid tools and how many images each monthly plan makes are covered in the image generation AI tools comparison.
2. 10 fun AI drawing ideas: tools, steps, and what to type
The games get a little more involved as you go down. Words only games can start right away; for games marked Uses a photo, read the cautions in section 3 first. Replace the [brackets] in each prompt with your own words.
① Turn a doodle into artwork
- Photograph the drawing straight from above in a bright spot (shadows get mistaken for lines)
- Attach the photo and name the style you want
- If the shapes change too much, add "do not change the shape of the lines"
Prompt Keep the shape of the lines in this doodle exactly as they are and turn it into a [soft-colored watercolor] illustration. Use a white background.
There is also a way to play without a photo. In AutoDraw, you draw on the screen with a finger or mouse, and it guesses what you are drawing and lines up illustration suggestions. Guessing "what did the AI think this was?" from a child's lines is a fun game.
② Redraw a photo in a new style
- Choose a photo with no human faces, such as travel scenery, a pet, or food
- Ask for the style together with "keep the composition"
- Make 3 or 4 versions of the same photo, changing only the style, and line them up
Prompt Keep the composition of this photo and turn it into [an oil painting with bold brushstrokes / pixel art / a paper cutout].
Ask for a style by the name of a technique, such as "oil painting," "pixel art," or "paper cutout." Avoid asking by the name of an artist or studio that is working today (section 3).
③ Create your own character and send it on a trip
- Decide on five visual features and make the first picture (color, shape, clothes, an item, expression)
- In the same conversation, say "the same character" and change only the scene
- If the face starts to drift, attach the first image and ask again
Prompt Draw a robot with [a red scarf, a round body, blue eyes, an umbrella in one hand, and a constant smile]. Next, put the same character on [a rainy street corner].
Keeping one character consistent across many pictures is one of AI's weak spots. OpenAI also lists keeping recurring characters consistent as something it may struggle with in its image generation limitations. Describing the features with the same words every time helps the character hold together.
④ Turn a child's drawing into a storybook page
- An adult, using their own account, photographs only the child's drawing (no face, name, or name tag)
- Ask to add a background and color while keeping the main figure
- Let the child decide the story and make a few more pages
Prompt Keep the shape and colors of the [dinosaur] in this drawing and turn it into a bright storybook page. The background is [a forest with a volcano].
An adult does the operating (section 4). Combine it with the "story starring your child" idea in AI activities for kids to make a whole picture book.
⑤ Original stickers
- Make four expressions of the same character
- Save the ones you like and bring them into Canva
- Add words like "thank you" later in Canva
Prompt Four versions of the same [Shiba Inu] character: smiling, surprised, crying, and angry. Sticker style with a white outline, on a white background.
Text inside images tends to come out garbled, especially longer phrases and non-Latin scripts. It is more reliable to have the AI make only the picture and add the words in Canva.
⑥ Four-panel comics and storybook illustrations
- Split the story into four scenes and write them out first
- Put all four panels on one image, or make them one at a time
- Make them without dialogue and add the words afterwards
Prompt A four-panel comic in one image. 1: the cat hero oversleeps. 2: runs in a panic. 3: slips in a puddle. 4: finds a rainbow and laughs. Same art style, no dialogue.
To build the story itself with AI, an AI text adventure works well. Adding illustrations to the best scenes makes it feel like a picture book.
⑦ "What if" mashups
- Combine two things (avocado × sofa, cloud × house)
- Decide how to present it, such as "like a furniture ad photo"
- Put the same words into another tool and compare how each draws it
Prompt A sofa that looks exactly like an avocado, with the pit as a cushion. Styled like a furniture catalog photo.
Entering the same words into several tools shows the differences in color and presentation. Firefly Free lets you pick models from several companies on one screen, which makes it good for comparing (non-Adobe models require being 18 or older).
⑧ Homemade coloring pages
- Always include these three: "black-and-white line art," "no filled areas," and "thick lines"
- Ask for portrait orientation to fit the paper
- Print it and color with colored pencils or crayons
Prompt A coloring page for kids. Theme: [dinosaurs and a volcano]. Black-and-white line art, no filled areas, thick lines, not too detailed, portrait orientation.
Gray shading makes pages hard to color. If it appears, add "no shading and no gray."
⑨ Room or garden makeover simulation
- Photograph your own room with no people in it (also leave out the view from the window, mail, and nameplates)
- Say what to change and what to keep
- Line up 2 or 3 ideas and choose with your family
Prompt Keep this room's layout and window positions and redo it in a Scandinavian style based on wood grain and white. Make the sofa [muted green].
The result is an impression. Measure furniture sizes and whether things will really fit separately.
⑩ Prompt lottery and drawing duel
- Ask the AI for a drawing prompt
- People draw the same prompt on paper or in AutoDraw, and the AI draws it as an image
- You can also flip it into a quiz: look at a picture the AI made and guess what words were used
Prompt Give us one easy and one hard prompt for a drawing duel. Don't show your own picture yet.
If you run out of ideas, try the fun AI prompts to try.
3. Before games with photos: faces, other people's photos, copyright
Games ①②④⑨ hand one of your photos to the AI. Before you do, look at everything in the picture.
Choosing a photo
- Photos you can use
- Yourself, your pets, your room, scenery, food, and photos the people in them have agreed to.
- Photos to avoid
- Photos of friends, classmates, celebrities, or anyone who has not given permission. Photos of children's faces. Photos that reveal a school name, address, or name tag.
- Requests to avoid
- Asking for existing characters, or copying the style of an artist or studio working today by name.
Each company's terms also set rules about other people's photos and rights. Here are the parts that matter for these games, taken from the original text.
| Service | What the terms prohibit (parts relevant to these games) |
|---|---|
| ChatGPT | Using someone's likeness, including their photorealistic image or voice, without their consent in ways that could confuse authenticity (OpenAI Usage Policies) |
| Gemini | Violating the rights of others, including privacy and intellectual property rights; impersonating a real person without explicit disclosure in order to deceive; non-consensual intimate imagery (Generative AI Prohibited Use Policy) |
| Copilot | Creating or editing images, voice, or video of other people (deepfakes) without their permission; infringing intellectual property and publicity rights (Copilot Terms of Use) |
| Canva | Content that invades privacy or publicity rights or infringes intellectual property (Acceptable Use Policy); misleading anyone that AI-generated content is human-made (AI Product Terms) |
| Firefly | Entering prompts designed to generate copyrighted, trademarked, or otherwise infringing content; violating privacy or publicity rights; impersonating other people (Generative AI User Guidelines) |
When you show a picture to others, add "made with AI." This matters most for pictures that could be mistaken for real photos, such as a real place with something in it that does not exist. Images made in ChatGPT or the Gemini app carry an invisible mark showing they were made with AI. The kinds of marks, and how to think about copyright and commercial use of what you make, are covered in section 7 of Getting started with AI image generation.
4. Age requirements when playing with kids
Drawing tools also set an age for having your own account. Rearranging the table in section 1 by who operates the tool gives the following.
- Under 13
An adult operates the tool from their own account, and the child sits alongside and thinks up the words. ChatGPT and Copilot are for 13 and older, and the only way for children under 13 to use Canva is Canva Education for schools. Photograph only the child's drawing, never their face. For children who love drawing by hand, trying AutoDraw together with an adult is a good option.
- 13 to 17
Words-only games can be played on the teen's own account in ChatGPT (with permission from a parent or legal guardian), the Gemini app, and Copilot. Transforming photos in the Gemini app requires being 18 or older, and so does using non-Adobe models in Firefly. Copilot's terms say some features may be turned off or limited for users under 18. In ChatGPT, parental controls can turn off image generation.
- 18 and older
Every game in this guide can be played on your own account. Even when uploading a child's photo to your own account, do not use photos that show their face.
These are only the conditions that matter for drawing. Steps for parental controls, age differences by country, and example family rules are in 10 AI activities to enjoy with kids, together with each company's original text.
5. How to get closer to the picture you want
When the picture is not what you pictured, the words are usually what is missing. Adding these four gets you closer.
Four things to add
- Subject
- "A smiling Shiba Inu puppy" instead of "a dog." Make the main subject clear.
- Style
- One word naming the technique: watercolor, oil painting, photo-style, pixel art.
- Mood and color
- "At dusk," "in pastel colors," and so on.
- Fixes
- Ask for one change at a time: "keep everything the same, but make only the background white."
AI struggles with exact text inside images, placing things exactly where you want them, and keeping one character consistent across many pictures (OpenAI's image generation limitations). When playing, don't worry about it too much, and add text afterwards. Prompt patterns and how each tool expects to be asked are in Getting started with AI image generation.
Summary
- Pick one tool from the free versions of ChatGPT, the Gemini app, and Copilot. Finish in Canva, and use AutoDraw for hand-drawn lines.
- The 10 games come in two types: words only, and games with photos. For photo games, look at what is in the picture first.
- Only use yourself, your pets, people who said yes, and your own room. Don't copy existing characters or artists by name.
- For children under 13, an adult operates the tool from an adult account. Transforming photos in the Gemini app requires being 18 or older.
- When the picture is off, add the subject, style, and mood, then fix one thing at a time.
For more ways to play, see 15 fun ways to use AI; to combine pictures with a story, try an AI text adventure.
FAQ
Q. How much can I do for free?
All 10 games in this guide work on free versions. ChatGPT, the Gemini app, and Copilot can make images for free, and Canva and Adobe Firefly also have free allowances. Every company says only that there is a limit, without giving a number, and the Gemini app's limit is daily, so if you get stopped, switch to another tool or wait until the next day.
Q. Can I use photos of my child?
As a rule, don't use photos that show their face. The games that use a child's drawing (① and ④) work fine if you photograph only the drawing. Transforming photos in the Gemini app requires being 18 or older, so it can't be done on a child's own account.
Q. Can I turn a friend's photo into an anime-style picture?
Not without their permission. Copilot's terms forbid creating or editing images of other people without their permission, and the terms of Google, Adobe, and Canva also forbid violating privacy or publicity rights. Even with permission, agree with the person on where the picture will be posted.
Q. Can I post or sell what I make?
When you post it, add "made with AI." To sell it or use it for work, check the terms of the tool and plan you used. Avoid pictures made to look like existing characters or a particular artist's style. How commercial terms compare is in section 6 of the image generation AI tools comparison.
Q. Which tool should I start with?
If you already use a chat AI, that is enough. If not, start with ChatGPT or the Gemini app, which are easy to refine through conversation. To keep kids drawing with their own hands, pair paper with AutoDraw so the game doesn't rely entirely on AI.