Coffee Making AI Prompt: Two Photos, Roles and Timing

Do you need to write a prompt on this website?

No. Our generator supplies a prepared reference and role instructions. You provide two separate photos, confirm permission and choose 480p or 720p for a 15-second video. The example below is for learning or adapting a prompt in a tool that accepts reference images; it is not an editable setting in this generator.

Start with the two photo roles

Photo 1 controls the main character on the right, with the spoon and red mug. Photo 2 controls the partner on the left, with the kettle and milk. Describe each subject separately. Keep species, coat markings or clothing tied to its own image, and use one clear subject per photo.

Our published example uses a Golden Retriever and a French Bulldog. It establishes one accepted 15-second 720p result. Cats, couples and other character combinations have not been validated by that sample.

A reusable two-photo coffee-making prompt

Create one continuous 15-second vertical kitchen-comedy shot from the two reference images. Keep the subject from photo 1 on the right with the spoon and red mug, and the subject from photo 2 on the left with the kettle and milk. Preserve each subject separately: face, species, coat pattern or clothing, body shape and natural anatomy. Show the coffee grounds missing the mug, the water pour, the mug being lifted, the milk spilling onto the counter and the final calm reaction. Keep a fixed camera and consistent lighting. Do not merge the characters, exchange their roles or add extra subjects. Keep animal paws rather than human hands. No captions or logos.

This is an example scene instruction, not a promise that any model will reproduce the exact motion. A tool that accepts only text may invent the characters instead of following two supplied photos.

Plan the whole 15-second sequence

Use the beginning to establish both characters and the spoon. Keep the coffee and water actions in the middle. Leave enough time for the mug lift, milk spill and final reaction. Avoid overloading a short clip with extra camera moves, transformations or a second location.

The reference and the model influence the final timing. Our accepted result has some pouring differences from the reference, so inspect the complete video before posting.

Treat music as a separate delivery choice

Our workflow adds fixed music after generation and also prepares a genuinely silent download from the same video. Neither download starts another generation. For a different song, use the silent file and add a sound you may use on the publishing platform.

There is no need to assume every coffee meme uses one universal song. Creator references and sound labels do not themselves establish permission for commercial reuse.

Fix unclear photos before adding more instructions

Use one well-lit subject in each JPEG, PNG or WebP image, up to 10 MB per image. Keep faces and distinguishing details visible. For dogs, include paws and body shape where possible. A longer prompt will not reliably recover details missing from a dark, blurred or tightly cropped photo.

Check both identities, paws or hands, mug contact, pouring and the ending. A higher resolution cannot guarantee correct anatomy or movement.

Use the prepared workflow and check the price

The one-time packs are $9.90 for one 15-second 480p video, $14.90 for one 15-second 720p video, or $34.90 for three 15-second 720p videos. There is no free generation or subscription. A confirmed failed task returns its credits; cash refund conditions are explained in the refund policy.

Use the related two-photo guide to view the original dog photos and real result, or open the generator from the homepage when you are ready.