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 ByteDance Seedream 5 Pro 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
ByteDance image model on Lovart ME

Seedream 5 Pro AI Image Generator Generate a new image or direct a reference-based edit

Write a visual brief or upload reference images, then choose one of eight aspect ratios and render at 1K or 2K. The same workspace handles original generation and controlled edits, with PNG or JPEG delivery.

2K
Maximum output
Choose 1K or 2K
8
Aspect ratios
Square, portrait, landscape, and ultrawide
4
Reference slots
JPEG, PNG, or WebP in this workflow
2
File formats
Export PNG or JPEG

Where Seedream 5 Pro fits in an image workflow

The useful distinction is not a vague quality label. It is the ability to start from text or combine several visual references while keeping the same ratio, resolution, and delivery controls.

Original image generation

Describe the deliverable, subject, scene, composition, lighting, and exclusions. Use a concrete brief so the first result can be reviewed against visible requirements.

Reference-led edits

Upload an existing image and state what must change and what must remain. Separate identity, layout, material, lighting, and copy instructions instead of using a broad request such as make it better.

Multi-image direction

Use separate references for a product, person, setting, palette, or treatment. Name the role of each reference in the prompt so conflicting visual signals are easier to resolve.

Format-ready composition

Set the destination ratio before prompting. A vertical cover, square catalog tile, presentation visual, and ultrawide banner need different subject placement and copy space.

Choose resolution and format for the stage of work

Seedream 5 Pro exposes two quality levels through this API. Basic produces 1K and High produces 2K. The live credit count changes with the selected resolution.

1K for working drafts

Use 1K to compare prompts, framing, color direction, and reference combinations. It costs 7 credits per generated image in the current integration.

2K for approved directions

Move to 2K after the composition works. It costs 14 credits and gives more room to inspect edges, faces, product details, and embedded text.

A practical map for the eight ratios

01

Square and balanced

Use 1:1 for catalog tiles, profile posts, icon-like compositions, and centered product images.

02

Portrait and mobile

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

03

Landscape and editorial

Use 3:2, 4:3, or 16:9 for editorial images, slides, web headers, and scene-based compositions.

04

Ultrawide placement

Use 21:9 for banners and cinematic backdrops. Tell the model where the subject and clean copy area should sit.

Seedream 5 Pro prompt templates for production tasks

Replace the bracketed fields and remove instructions that do not affect the deliverable. For edits, identify each reference by order and assign it one clear job.

Reference-based scene change

Move a subject into a new setting without losing the required visual traits.

Prompt Edit reference 1. Keep the subject identity, pose, clothing construction, and camera angle. Replace only the background with the location and atmosphere from reference 2. Match contact shadows and reflected light. Remove [unwanted object]. Do not add people or text.

Working note: List preservation rules before the requested changes.

Editorial key visual

Create an image that can carry a headline and supporting layout.

Prompt Create a [ratio] editorial key visual about [topic]. Main subject: [subject and action]. Composition: [framing], with clear copy space in the [area]. Palette: [colors]. Lighting: [description]. Keep symbols literal and relevant; no decorative lettering or watermark.

Working note: Specify the empty area as deliberately as the subject.

Material and color variation

Test a product variation while keeping geometry stable.

Prompt Using reference 1 as the fixed product geometry, change the finish to [material] and the colorway to [colors from reference 2]. Preserve dimensions, seams, controls, logo position, and viewing angle. Use a neutral studio background and realistic material response.

Working note: Review seams, reflections, contact points, and small controls at full size.

How to use Seedream 5 Pro on Lovart ME

Treat each generation as a short production brief. The controls make it easier to change one variable at a time.

  1. 01

    Choose text or image mode

    Use Text to image for a new concept. Choose Image to image when identity, geometry, composition, palette, or another visible property must come from a source image.

  2. 02

    Write the brief and assign references

    State the deliverable first. Add subject, setting, composition, lighting, and exclusions. When using several images, explain what each one controls.

  3. 03

    Set ratio, resolution, and format

    Match the ratio to the publishing slot. Use 1K while testing and 2K for an approved direction. Choose PNG for lossless graphics or JPEG for a smaller photographic file.

  4. 04

    Generate and inspect

    Check identity, hands, product geometry, small text, edges, shadows, and unwanted additions. Revise one or two constraints at a time.

Seedream 5 Pro questions

These answers reflect the documented Kie.ai parameters and the controls available in this Lovart ME integration.

What is Seedream 5 Pro?

Seedream 5 Pro is a ByteDance image model available through separate Kie.ai text-to-image and image-to-image endpoints. This page combines both workflows in one generator.

Can Seedream 5 Pro edit an existing image?

Yes. Switch to Image to image, upload at least one reference, and describe both the required changes and the details that must remain fixed.

Which resolutions are available?

The documented API provides Basic quality at 1K and High quality at 2K. Lovart ME shows 7 credits for 1K and 14 credits for 2K before generation.

Which aspect ratios can I use?

The available ratios are 1:1, 4:3, 3:4, 16:9, 9:16, 2:3, 3:2, and 21:9.

Which input and output formats are supported?

Reference uploads accept JPEG, PNG, and WebP files up to 10MB each. Generated images can be requested as PNG or JPEG.

Does this generator include layer decomposition?

No. The pricing list identifies layer decomposition as a separate operation. This page integrates only the documented text-to-image and image-to-image endpoints.

How should I use multiple reference images?

Give each reference a specific role, such as product geometry, subject identity, location, palette, or material. Remove references that do not resolve a clear production decision.

Can I use the result commercially?

Usage depends on Lovart ME terms, the rights attached to your references, and local law. Confirm you can use uploaded material and review the current terms before commercial publication.