AI Image Generator
Models
Nano

Google's cutting-edge image generation model

3 credits
Fast
High Quality
Gpt

OpenAI's GPT-4o image generation model

3 credits
High Quality
Text & Image
Seedream

ByteDance image generation and reference-based editing in 2K, 3K, or 4K

6 credits
2K/3K/4K
Text & Image
Flux

Flux Kontext standard model with balanced performance

4 credits
Fast
High Quality

Powered by Google Nano Banana 2 Lite Official API

Public Visibility

When this option is enabled, the output image may be selected by AI Generate and published to the Explore.

Notification
This is a notification message
Sample Images
Gemini 3.1 Flash Lite Image on Lovart ME

Nano Banana 2 Lite AI Image Generator Fast 1K drafting and editing for repeatable visual work

Turn a written brief or up to 10 reference images into a 1K visual. Nano Banana 2 Lite costs 3 credits per image on Lovart ME and offers 15 aspect ratios for quick concept tests, edits, and campaign variations.

3
Credits per image
Fixed cost for one 1K result
10
Reference images
JPEG, PNG, or WebP inputs
15
Aspect ratios
Auto plus standard and extreme formats
1K
Output size
Built for fast drafting and iteration

Where Nano Banana 2 Lite fits

Google describes Gemini 3.1 Flash Lite Image as its fastest, most cost-efficient Nano Banana image model. It is a practical choice when iteration speed and output volume matter more than maximum resolution.

Rapid concept drafts

Compare compositions, palettes, camera angles, and art directions at 1K before committing more credits to a higher-resolution model.

Reference-led edits

Upload product shots, characters, layouts, or style cues and state what each reference controls. The clearer the role of each image, the easier the result is to review.

Text inside images

Draft posters, cards, labels, and social graphics with short copy. Put exact wording in quotation marks and proofread every character before publishing.

Consistent visual subjects

Reuse the same reference order and repeat defining details such as clothing, materials, colors, and facial traits when building a sequence of related images.

High-volume direction testing

Use the fixed 1K output for ad variants, thumbnail options, product scene ideas, and other workflows where many directions need a quick first review.

Plan around a fixed 1K output

Nano Banana 2 Lite on Lovart ME produces one 1K image per job. Use it to settle the visual direction, then move an approved concept to Nano Banana 2 or Nano Banana Pro when the final asset needs more resolution.

Draft the full composition

Describe the final placement, subject scale, copy area, and background from the start. A complete 1K draft is more useful than a detailed crop with no room for the deliverable.

Inspect production details

Review faces, hands, product geometry, logos, small objects, and all generated text at full size. Regenerate anything that will not survive a real publishing workflow.

Escalate only approved work

When a concept is accepted, carry its prompt, reference order, and ratio into a higher-resolution model. This keeps expensive iterations focused on a proven direction.

Choose the ratio for the destination

01

Square and product

1:1, 4:5, and 5:4 work well for catalog tiles, profile posts, and product compositions with balanced margins.

02

Portrait and mobile

2:3, 3:4, and 9:16 suit people, posters, covers, and full-screen mobile placements.

03

Landscape and campaign

3:2, 4:3, 16:9, and 21:9 fit editorial images, presentations, web banners, and cinematic scenes.

04

Extreme formats

1:4, 4:1, 1:8, and 8:1 need explicit subject placement and copy space. Use Auto only when the source image should guide the crop.

Nano Banana 2 Lite prompt templates for working drafts

Replace the bracketed fields and keep requirements in priority order. These templates are built for reviewable outputs, not decorative prompt length.

Social campaign variants

Produce a controlled set of visual directions for one message or offer.

Prompt Create a [ratio] social image for [campaign]. Main subject: [subject]. Message hierarchy: [headline area, subject, supporting detail]. Explore [style direction] with [palette]. Keep the logo area clear and do not add unapproved words, dates, or prices.

Working note: Change one variable at a time, such as the camera angle or palette, so reviewers can compare the results.

Reference image edit

Change a defined part of an image while keeping the rest stable.

Prompt Edit reference 1. Keep [subject, framing, colors, and details] unchanged. Replace only [target area] with [new content]. Match the existing perspective, light direction, shadows, and material texture. Do not crop the image or add new objects.

Working note: List the protected details before the requested change to make the boundary easy to follow.

Short-copy poster

Draft a poster or card with a small amount of text inside the image.

Prompt Design a [ratio] poster for [event or offer]. Show the exact headline '[headline]' and line '[supporting line]'. Use [visual style], strong contrast, and a simple reading order. Do not invent sponsors, dates, logos, or additional text.

Working note: Generate one language at a time and proofread all copy before the image leaves review.

How to use Nano Banana 2 Lite on Lovart ME

A short setup keeps fast generation useful and makes revisions easier to compare.

  1. 01

    Choose text or image mode

    Start with Text to image for a new concept. Choose Image to image when an existing product, person, layout, or style must guide the result.

  2. 02

    Write the deliverable first

    Name the asset, audience, subject, composition, and constraints. If you upload several references, assign a clear role to each one.

  3. 03

    Set the publishing ratio

    Select a ratio that matches the destination instead of planning to crop later. Use Auto when preserving the source image composition is more important.

  4. 04

    Generate and review at full size

    Check identity, product shape, hands, typography, edges, and small objects. Revise one or two instructions per attempt so changes remain traceable.

Nano Banana 2 Lite questions

These answers separate Google model information from the options currently available through Lovart ME.

What is Nano Banana 2 Lite?

Nano Banana 2 Lite is Google's Gemini 3.1 Flash Lite Image model. Google positions it as the fastest, most cost-efficient image model in the Nano Banana family, aimed at rapid ideation and high-volume workflows.

How is Nano Banana 2 Lite different from Nano Banana 2?

Lite prioritizes speed, throughput, and lower generation cost with fixed 1K output on Lovart ME. Nano Banana 2 is the general-purpose option and supports 1K, 2K, and 4K output here, so it is a better fit when the approved image needs more detail.

How many credits does one image cost?

Each Nano Banana 2 Lite job currently costs 3 Lovart ME credits and returns one 1K image. The generator shows the cost before submission.

How many reference images can I upload?

You can upload up to 10 JPEG, PNG, or WebP reference images. Lovart ME currently limits each upload to 10MB. Use fewer references when possible and explain the role of every image in the prompt.

Which aspect ratios are available?

The generator supports Auto plus 1:1, 1:4, 1:8, 2:3, 3:2, 3:4, 4:1, 4:3, 4:5, 5:4, 8:1, 9:16, 16:9, and 21:9.

Can Nano Banana 2 Lite generate accurate text?

Google reports legible in-image text rendering, but generated copy still needs proofreading. Keep the wording short, quote the exact text, and avoid asking the model to invent legal, pricing, or factual information.

How fast is Nano Banana 2 Lite?

Google reports a 4-second text-to-image benchmark in its own environment. Lovart ME does not guarantee a fixed completion time because queue load, network conditions, and the upstream service can affect each job.

Can I use generated images commercially?

Commercial use depends on current platform terms, your rights to uploaded material, and the laws that apply to your project. Review the output and confirm those rights before publication.