Add your character’s views
Drop, paste or choose up to six images: front, side, three-quarter, back, close-up and full body. Each view gets a label you can change.
This character reference sheet maker turns your front, side, back and close-up images into one labelled sheet with name, outfit, hair, colour palette and a reusable prompt anchor.
Use the PNG as the reference image for AI short dramas, and keep the JSON for your prompts.
Drop up to 6 views of your character here, or paste them
Front, side, three-quarter, back, close-up, full body · JPG, PNG, WebP, AVIF · Nothing is uploadedDrop, paste or choose up to six images: front, side, three-quarter, back, close-up and full body. Each view gets a label you can change.
Pick a layout, fill in name, age, role, outfit and hair, check the palette taken from the images and write the prompt anchor and the avoid list.
Download the sheet as a 2048 px PNG to use as a reference image, and the JSON to keep the text for every prompt.
Seedance 2.5, Wan 3.0 and Kling 4.0 on WaveSpeed AI each take reference images next to your prompt and can make vertical 9:16 video for short dramas. Upload the sheet as a reference image, name it in the prompt the way the model expects, and paste your prompt anchor.
Text-to-video with reference images · Add the sheet under reference images and describe the scene.
In the prompt, call the sheet: @Image 1
Reference-to-video · Add the sheet as one of up to 10 reference images.
In the prompt, call the sheet: Image 1
Reference-to-video · Add the sheet as one of up to 10 reference images.
In the prompt, call the sheet: @Image1
Each model’s page lists everything it accepts, its current price and examples.
No. Laying out, labelling, picking colours and exporting all happen in your browser. Your images never leave your device, and the tool works without an account.
One image that shows a character from several angles with their name, outfit, hair and colours written next to them. Artists use it to keep a character consistent; AI video models use it as a reference image for the same reason.
Front, side and back show the whole outfit; a three-quarter view and a close-up show the face. For AI video, a clear front view and a close-up help the most. You can add up to six images and label each one.
A short paragraph that describes the character the same way every time. Paste it into each prompt together with the sheet so every shot starts from the same description. Draft from fields writes a first version from what you filled in.
From images takes the main colours of your pictures and leaves out a plain backdrop. Pick from sheet adds the colour you click on the preview, and Add colour lets you choose any colour. Click a swatch to change it.
The name, age, role, outfit, hair, palette, prompt anchor, avoid list (also as one line for negative prompt fields), the layout and the view labels. Open a saved JSON to bring the text back and keep editing.
2048 pixels on the long side: 2048 × 1152 for 1 × 3, 2048 × 1536 for 2 × 3 and 2048 × 1792 for hero + details.
Generate the shots with Seedance 2.5, Wan 3.0 or Kling 4.0 in the video generator, or plan scenes and recurring characters in Cinema Studio and Influencer Studio.