Yeri AI

Yeri AI is a free web workspace for making images and short video drafts. You open https://yeri.ai in a browser. There is no desktop client to install. The same workspace covers text to image, image to image, text to video, and image to video. It also includes an AI image editor, photo effects, templates, a background remover, and an upscaler. The models you can use there are Seedream, Nano Banana, GPT Image, Kling, and Seedance. Seedream, Nano Banana, and GPT Image are the image side. Kling and Seedance are the video side. You stay on one page while you switch between those jobs. A practical way to work is to lock a small keyword set before you generate anything. Write one phrase for the subject, one for the scene, one for the light, and one for the style. A first prompt can be as plain as a ceramic bottle on a pale desk, soft window light, editorial photo. Generate a few frames. Keep the one whose composition is readable, whose subject is complete, and whose color can sit on a brand board. If it is not right, change only one of those four phrases and run again. Do not rewrite the whole prompt on every pass. Image to image is for the moment you already have a frame that is close. Keep the composition and change the material, the background, or the season. Background removal and upscaling come after that choice, not before. Running a background remover or an upscaler on a frame you will throw away wastes the step. Photo effects and templates are the fast path when a cover is due. A template gives the layout. You swap in your own subject. Move to video only after the still is stable. Text to video and image to video here are for a short check, about six to ten seconds, not a finished film. Watch whether the camera move hides the subject and whether the pace holds. Compare Kling and Seedance on the same prompt if you want to see which motion stays steadier. Keep the winning frame and the original prompt in one note so someone else can repeat it. This fits product concept shots, social covers, and storyboard drafts. It is not a pipeline that publishes without a person looking. Someone still picks the frame. Brand work should keep the same style words and save the prompt with the export. Start at https://yeri.ai.