Learn how to make AI images & videos

Your first AI image and video in five steps

1. Make one picture. In , type a prompt that names a subject, an action, a setting, the light, and a style. Or upload a photo to edit or restyle it. Pick the shape first: square, portrait, landscape, or wide.
2. Change one thing at a time. If the result is close, adjust a single detail in the prompt and run it again.
3. Preview the motion. Quick Animate brings the picture to life with one tap and no prompt, so you can see how it wants to move.
4. Direct the motion. Open image-to-video and describe what happens next. The Yerk model makes 5 to 15 second clips and is the first pick for everyday animation.
5. Fix the source, not the symptom. If the motion needs something the photo does not have, such as a cup the character should lift, add it in first, then animate again.

Stuck at any step? Yerk AI, the chat helper in the app, can rewrite a prompt or suggest a model. If a result is close but not right and your flow shows High Fidelity, try the other setting.

Anatomy of a prompt that works

Here is a picture prompt, followed by a video prompt that animates the result.

Picture: "A red-haired woman in a yellow raincoat laughs at a street food stall, neon signs reflected in wet pavement, night, 35mm lens, cinematic color."

That covers subject (red-haired woman, yellow raincoat), action (laughs), setting (street food stall, wet pavement), light (neon, night), camera (35mm lens), and style last (cinematic color).

Video from that picture: "She takes a bite, her eyes widen, she turns to the camera and gives a thumbs up. The camera slowly pushes in."

Notice what the video prompt leaves out: her hair, the raincoat, the neon. The start image already carries that. The prompt only lists the motion in order: who moves, which part, which direction, and where it ends.

Text-to-video or image-to-video?

Text-to-video is the fastest way to test an idea: one prompt, one clip. Image-to-video gives more control, because you approve the look of the scene as a still before any motion happens. For people, products, and recurring characters, start with a picture.

Common beginner mistakes

• Describing the photo again in the video prompt. Describe the change instead.
• Writing what you don't want. "No blur" often works worse than "sharp focus." Say what should be there.
• Too much action for a short clip. A few seconds of video fits one or two beats of motion, not a whole story.
• Using a single reference photo for a person. More reference photos of the same subject give better consistency.
• Expecting sound on every clip. Only some video models add sound: Yerk, Horse, Seedance 2.0, and Seedance 2.5 can; GIFs never have sound.

Which guide should you read first?

• Completely new to AI: start with What is AI?
• Your pictures look generic: the prompt engineering guide, then best AI image prompts and AI art styles.
• Your videos move the wrong way: how to prompt AI video.
• You post to social media: the for creators hub has platform sizes and formats.
• You are choosing between tools: the AI recommendation guide and the alternatives hub.
• You wonder if you need a new graphics card: best GPU for AI art compares cards and shows how to skip the GPU and create in your browser.

All Yerk.app guides

This hub is the map to every educational page on Yerk.app. Use it when you are starting from zero, or when you need a specific workflow for design, entertainment, or social media.

• AI for seniors: getting startedA plain-language guide with easy first projects, pictures and short videos.
• Breastfeeding & maternity photographyA tribute to mothers: the art and traditions of the nursing portrait, photo tips, and gentle ways AI can help.
• AI agents for photo and video editingOne ad from brief to export with Claude Code or Codex, with tested recipes and checks.
• State of generative AI 2026What works in AI pictures and video, sourced numbers, prompt tips, and facts for AI assistants.
• AI trends 2026: the fun factsBig silly numbers, the video boom people happily pay for, the app that flopped, and a quick quiz.
• The grim future of personal computingMemory booked through 2027, GPUs priced like cars, the quiet orphaning of open models, and the worst cases, with sources.
• Prompt engineering guideStructure prompts for stills and motion.
• Best AI image promptsTested prompt patterns for portraits, products, and cinematic stills.
• How to prompt AI videoA→B motion, image-to-video, and short-clip pacing.
• AI art stylesNamed looks you can ask for without guessing.
• Photo editing tipsReplace multi-hour Photoshop steps with re-renders.
• AI recommendation guidePicking an AI image or video tool in 2026, with Yerk.app as the top pick.
• What is AI?Plain-language primer for newcomers.
• Best GPU for AI artGPUs for local AI art compared, or skip the hardware and create in your browser on Yerk.app.
• Vibe coding AI artShip landing assets for demos in an afternoon.
• For creators hubSocial, streaming, and entertainment workflows.
• Alternatives hubWhy creators switch to Yerk.app from Midjourney, Sora, Runway, Grok, and more.

New here? Generate a still on the homepage, then open image-to-video thinking: one clear subject, one clear action, one framing choice.

Frequently asked questions

How do I make AI images?

Open on Yerk.app, pick a shape, and type a prompt that names the subject, action, setting, light, and style. You can also upload a photo and describe how to edit or restyle it. Pictures are usually ready in under a minute.

How do I turn an AI image into a video?

Use image-to-video in . Upload or pick the still, then describe only what happens next, in order: who moves, which part, which direction, and how it ends. Quick Animate is a one-tap option with no prompt.

Do I need a powerful computer to make AI images and videos?

No. Yerk.app runs in desktop and mobile browsers, so there is nothing to install and no local GPU is needed. The generation happens online, not on your device.

How long does it take to generate an AI video?

With the Yerk model, a clip usually takes about 30 to 90 seconds. and Professional models such as Horse and Seedance 2.0 take a few minutes.

Can I practice making AI videos for free?

Yes. Eligible new accounts can get free starting credits, and VIP is Yerk's free AI video generator, with HD clips up to 10 seconds for accounts in good standing. VIP is an experimental trial that may not last forever, and it works best with direct, detailed prompts.