Image, video, and e-commerce image guide

Who this site is for

• Design and art teams: Quickly validate posters, e-commerce images, brand visuals, and UI concepts.
• E-commerce and operations teams: Reuse product pages, hero images, campaign posters, ads, and image-set templates.
• Content creators: Create infographics, storyboards, covers, social posts, and short-video assets.
• Short-video and advertising teams: Use Grok Imagine 1.5 for product showcases, ad shots, and vertical clips.
• Prompt users: Compare structured prompts across different scenarios.
• Developers and product managers: Generate software UI concepts.

What is GPT-Image-2

GPT-Image-2 is an image generation and editing model for high-quality visual creation. It can generate images from text, edit reference images, render readable text, and create layouts such as posters, infographics, product pages, and software interfaces.

What is Nano Banana Pro

Nano Banana Pro is Google's image generation and editing model based on Gemini 3 Pro, also known as gemini-3-pro-image. It is designed for professional workflows, including text-to-image, reference-image editing, image fusion, and local revisions.

Its strengths include real-world knowledge, readable text and infographics, and detailed control over composition, aspect ratio, local edits, and multiple references.

What is Grok Imagine 1.5

Grok Imagine 1.5 is xAI's video generation model. CicPic supports text-to-video and up to seven reference images, producing 6-30 second videos for product showcases, advertising, social content, and brand campaigns.

It is useful for realistic camera movement, reference-image continuation, and short-video delivery. Supported ratios are 16:9, 9:16, 1:1, 3:2, and 2:3. Video results are stored in the current browser, so download successful results promptly.

What is VEO 3.1

VEO 3.1 is Google's video generation model for natural camera movement, realistic lighting, subject consistency, and commercial short-video storytelling. CicPic supports fixed 8-second landscape or portrait videos in 720p and 1080p.

What is an e-commerce image set

An e-commerce image set is a batch workflow for product operations. Upload a product reference, let the system reverse-engineer a consistent prompt set, and generate hero images, detail-page visuals, ads, selling-point graphics, and lifestyle scenes in one workflow.

Core capabilities

1. Text rendering

The models can render readable text for posters, charts, teaching graphics, menus, product selling points, and complete product detail pages. Put the exact text you need in the prompt.

2. World knowledge

The models understand many real-world brands, places, products, app layouts, historical subjects, and industry conventions. This makes them useful for interface screenshots, social profiles, game posters, and brand pages.

3. Precise editing

Upload a product photo or reference image and describe the intended change. You can replace a brand, change text, swap a person, preserve a scene, or produce a refined white-background product image.

4. Aesthetic range

Useful directions include cinematic stills, retro film, fashion photography, pixel art, comics, watercolor, and minimal technology. photorealistic is a useful keyword when a natural photographic result is desired.

5. Interface generation

The models can create social-media screenshots, app interfaces, e-commerce homepages, music players, game screens, desktop environments, and terminal-style layouts.

Prompt framework

[Task] + [Subject] + [Style] + [Technical parameters] + [Output specification]

Five elements

Principles

1. Specific beats vague.
2. You can write prompts directly in your preferred language.
3. Include the exact text that should appear in the image.
4. Use style references to anchor the visual direction.
5. Specify the aspect ratio and resolution.

Advanced tips

Tip 1: Use photorealistic

This keyword helps the model avoid a plastic appearance and preserve photographic details.

Tip 2: Use Thinking mode when needed

Use Thinking for brand knowledge, people, connected image series, or other tasks that need deeper reasoning. Instant mode is suitable for simple and fast generation.

Tip 3: Provide the text itself

Instead of asking for “a poster with a promotion,” provide the brand name, product name, price, and promotion copy exactly.

Tip 4: Anchor the aesthetic

Use directions such as “in the style of a 1960s French New Wave poster,” “pen and wash,” “modern Chinese luxury,” “pixel art,” or “retro film.”

Tip 5: Reference image plus editing

1. Upload a reference image.
2. Generate a first draft.
3. Use the image editor for targeted changes.

Tip 6: Use world knowledge

You can directly ask for an app homepage, a chat screenshot, or a product detail page without describing every familiar interface element.

Tip 7: Infographic structure

Use modular sections, rounded information blocks, clear heading hierarchy, a clean background, refined icons, and enough whitespace.

Tip 8: Choose a ratio

Video prompt framework

For Grok Imagine 1.5, describe the subject, scene, camera movement, action, lighting, duration, ratio, and quality.

[Subject] + [Scene] + [Camera movement] + [Action] + [Lighting] + [Ratio and quality]

E-commerce image-set prompts

First define the purpose of the image set, then give each image a distinct job:

Limitations

Physical 3D logic, dense textures, precise arrows and charts, facial consistency, and high-risk documents can still produce errors. Review results carefully and avoid sharing sensitive or misleading generated content.

Pricing

CicPic charges by model and resolution. Current examples include image generation from ¥0.1 per image, video generation from ¥0.1 per second, and e-commerce sets based on the selected image model and quality. Prices may change with third-party API costs, exchange rates, and service policies.

Disclaimer

Use generated media responsibly. Do not distribute forged identity documents, exam papers, or other high-risk content. People may not reliably distinguish generated media from real media, even when watermarked.