Nano Banana 2
Create cinematic stills, keyframes, and concept art in seconds
Image Generator Features
Everything you need to go from a text prompt or reference image to production-ready stills — fast, flexible, and built to feed into video.
Concurrent Generation
Run up to 5 image generations at the same time. Start new prompts while previous ones are still rendering — no waiting required.
Up to 4K Resolution
Generate images at 1K, 2K, or 4K resolution. Choose the level of detail that fits your use case, from quick drafts to print-ready stills.
Flexible Aspect Ratios
Choose from 1:1, 16:9, 9:16, 3:4, 4:3, 3:2, or match the aspect ratio of your input image automatically.
Image-to-Image Mode
Upload a reference photo, product shot, or rough concept and generate styled variations without starting from scratch.
Image-to-Video Pipeline
Use any generated still as a start frame for the video workflow. Approve the image first, then add motion — one unified pipeline.
Multiple Output Formats
Export in JPG, PNG, or WebP depending on your downstream workflow — landing pages, social posts, or print assets.
Batch Output
Generate up to 3 images per prompt in a single run. Compare directions side by side without re-entering the same prompt.
Two Models, One Workflow
Choose between Nano Banana 2 (up to 4K, cinematic stills) and Seedream 5 Lite (fast, cost-effective at 3 credits per image). Both support text-to-image and image-to-image.
Full Generation History
Every image you generate is saved to your workspace. Revisit, download, or send any past result into the video pipeline at any time.
Image Examples
Examples of the still directions teams typically generate before turning approved frames into campaigns, ads, or motion concepts.
Fashion campaign concepts
Generate polished first-look visuals for campaign reviews, moodboards, and social launch planning.
Editorial portrait directions
Use text prompts to align on framing, styling, and visual tone before production decisions are locked.
Product-ready still frames
Create clean reference images that can move directly into landing page concepts or video start-frame workflows.
Interior design concept
Useful for interior design teams reviewing spatial layouts, lighting, and material choices before finalizing directions.
Lifestyle scene composition
A good fit for product pages, launch decks, and concept reviews when you need environment and product emphasis to read clearly.
Creative illustration
Generate unique illustrative concepts that blend photorealism with artistic style for branding and creative projects.
Whimsical character art
Works well for creative reviews when the team needs playful, character-driven imagery for campaigns or merchandise.
Image FAQ
Practical questions teams usually ask before using image generation as part of their production workflow.
What is this page best for?
Use /image when you need a strong still first: concept frames, campaign visuals, moodboards, product imagery, or a start frame for the video workflow.
Can I upload a reference image?
Yes. Switch to Image to Image to upload a source image and generate cleaner, styled, or more production-ready variations.
Where do generated results appear?
The newest image appears in the preview panel, and your recent history stays available in the workspace. Every generation is saved automatically.
How should teams use image generation with video?
Most teams align on stills first, then take the approved frame into video generation so motion credits are spent on stronger directions.
Can this page help me prepare a start frame for video?
Yes. This is one of the main reasons the page exists. Generate a still here first, then carry that approved frame directly into the video workflow.
What kinds of prompts work best for images?
Prompts that specify subject, setting, mood, framing, lighting, materials, and use case usually produce stronger stills than short generic requests.
When should I use image-to-image instead of text-to-image?
Use image-to-image when you already have a rough frame, product photo, or creative draft and you want controlled variations without starting over.
How do teams usually review generated images?
Most teams compare a few still directions, pick the closest frame, and only then move into motion or page design work.
How many images can I generate at once?
You can run up to 5 concurrent generations and produce up to 3 images per prompt. Start new prompts while earlier ones are still rendering.
What resolutions are supported?
Nano Banana 2 supports 1K, 2K, and 4K output. Seedream 5 Lite generates at standard resolution with basic or high quality options. Higher settings use more credits.
What image formats can I export?
Generated images can be exported as JPG, PNG, or WebP. Choose the format that best fits your downstream workflow — web, social, or print.
Is there a way to generate images from an API?
Yes. Hula AI offers API access for programmatic image generation. Visit the API Keys section in settings to create your access credentials.
What is Seedream 5 Lite?
Seedream 5 Lite is a fast, cost-effective image model at just 3 credits per image. It supports both text-to-image and image-to-image modes with basic or high quality options — great for rapid iteration and high-volume work.
How do I choose between Nano Banana 2 and Seedream 5 Lite?
Use Nano Banana 2 when you need maximum resolution (up to 4K) and cinematic quality for hero images and print. Use Seedream 5 Lite when you need fast, affordable generations for exploration, batch work, or quick drafts.