Text-to-image and image-to-image both support transparent and semi-transparent backgrounds, so assets drop straight into composites with no cutout pass.

Native multilingual text

Small text, dense text, and multilingual copy render with correct letterforms, with tested support for Russian, Thai, Japanese, Filipino, and more.

4K rendering

Custom aspect ratios up to 8.29 megapixels, including 4K UHD (3840 × 2160) at 16:9, with sharper detail at the same pixel count.

Natural multi-image blending

Combines one or more reference assets into one scene with consistent scale, perspective, and lighting, while preserving subject and product identity.

Capabilities above are from ByteDance's own announcement for Seedream 5.5. The public API doc and pricing are listed as to be published; this page updates with the official figures the moment they do. Seedream 5.5 cannot be run yet, so every example image here was generated on Seedream 5.0 Pro, the current Seedream model on Morphic.

Technical specifications

ByteDance

Built by ByteDance, maker of the Seedream image line.

Announced

ByteDance has announced the capabilities and shared early docs. The public API doc and pricing are listed as to be published; it arrives on Morphic as soon as the API is live.

Image

Text-to-image and image-to-image, with transparent output supported in both.

4K · up to 8.29 MP

3840 × 2160 at 16:9, with flexible aspect ratios up to 8.29 megapixels. API presets of 1K, 1.5K, and 2K are convenience shortcuts, not the full limit.

Use cases

Two folded knit sweaters on a studio plinth, catalogue retouching that stays clean across edits

Multi-round retouching that holds up

Keep editing the same image across rounds for e-commerce, portraits, film, advertising, or games. Non-target regions stay pixel-identical and quality does not degrade.

Cut-out product and decorative assets laid out with clean isolated edges, ready to composite

Transparent assets for compositing

Generate logos, UI pieces, decorative text, and product cutouts with transparent or semi-transparent backgrounds, ready to layer in a design file.

A poster separating into background, image and text layers floating apart in depth

Layered posters and layouts

Split a finished design into up to 99 background, element, and text layers you can edit on their own, with structured text recognition for copy changes and localization.

A bakery shopfront whose sign and chalk menu render in both Latin and Japanese letterforms

Multilingual signage and marketing

Render accurate small text, dense text, and multilingual copy on product labels, posters, and infographics, across languages like Russian, Thai, Japanese, and Filipino.

A dropper bottle, a cupped hand and a towel blended into one scene under a single light

Multi-image product and portrait scenes

Blend one or several reference assets into one scene with convincing scale, perspective, and lighting, keeping subject and product identity intact.

A labelled cutaway diagram of a pour-over coffee brewer with small legible callouts

4K hero visuals and infographics

Render at up to 3840 × 2160 for high-density infographics, labelled diagrams, and hero imagery with small text that stays legible.

Prompt examples

Recolor

Change the kettle to glossy forest green, keep everything else exactly as it is

Edit prompt

Swap text

Turn the sign into one for a bakery called CRUMB, keep the style and lighting

Edit prompt

Split layers

Separate this poster into background, subject, and text layers

Edit prompt

Transparent asset

Generate a leafy watercolour wreath on a transparent background, PNG with alpha

Edit prompt

Multilingual text

Render the same shop sign in Japanese and in Thai, with correct letterforms

Edit prompt

Multi-image blend

Combine the chair, lamp, and rug references into one living-room scene with matched lighting

Edit prompt

Overview

Seedream 5.5 is ByteDance's next release in the Seedream image line, built around precise multi-round editing, transparent output, up to 99-layer separation, native multilingual text, and 4K rendering. ByteDance has published early documentation covering the capabilities below; the public API doc and pricing are listed as to be published.

What Seedream 5.5 brings

Multi-round editing is the headline. You can edit a region or element and every other pixel stays identical, and the result holds up across repeated rounds without quality degradation — the capability ByteDance highlights for e-commerce, portrait retouching, film and TV, advertising, and games.

Alongside it, text-to-image and image-to-image both support transparent and semi-transparent output, so logos, UI pieces, decorative text, and product cutouts drop into composites without a cutout pass. Layer separation now supports up to 99 reusable background, element, and text layers; text layers come back as a PNG plus the text itself, structured and readable, so downstream tools can edit the copy for localization or template reuse.

In-image text has been improved for small text, dense text, and multilingual copy, with ByteDance's own examples covering Russian, Thai, Japanese, and Filipino. Multi-image blending keeps subject and product identity when combining one or more reference assets, with consistent scale, perspective, and lighting. And rendering now reaches 4K: custom aspect ratios up to 8.29 megapixels, including 3840 × 2160 at 16:9. The API's 1K / 1.5K / 2K preset sizes are convenience shortcuts rather than the full limit.

What is still to come

ByteDance has not yet published the public API doc or pricing — both are listed as to be published, and a dated model id will land with the API release. This page updates with the official figures the moment they do.

Use it today

If you need a Seedream image model now, the current release is live on Morphic and runs in one workspace alongside the rest of the image line. Open the Seedream 5.5 image editor to start, and the Seedream 5.5 guide lays out what to expect in more detail.

Simple pricing

Get started for free today, with the option to upgrade or cancel anytime.

Basic

/ month

billed as $228 per year

2400 monthly credits

1 user only

All models

Workflows

Standard

/ month

billed as $288 per year

3625 monthly credits

1 user only

All models

Workflows

Pro

/ month

billed as $540 per year

6350 shared monthly credits

1 user

+ up to 4 more at extra cost

All models

Workflows

Max

/ month

billed as $2040 per year

24650 shared monthly credits

1 user

+ up to 9 more at extra cost

All models

Workflows

Enterprise

For higher limits

Custom

pricing and billing terms

High-volume credits

Custom seat limits

All models

Workflows

Free

For playing around

forever free

Up to 20 credits

1 user only

Limited models

Workflows

Compare plan details

FAQs

What is Seedream 5.5?

Seedream 5.5 is ByteDance's next release in the Seedream image line, announced with precise multi-round editing, transparent and semi-transparent output in both text-to-image and image-to-image, native multilingual text, up to 99-layer separation with text recognition, and 4K rendering.

When is Seedream 5.5 coming out?

ByteDance has announced the model and shared early documentation. The public API doc and pricing are listed as to be published; a dated model id lands with the API release. ByteDance announces its image models on its Seed blog, which is where the firm date lands first.

What can Seedream 5.5 do?

Multi-round region editing with non-target pixels preserved; transparent and semi-transparent output in text-to-image and image-to-image; up to 99 editable layers with structured text recognition; native rendering of small, dense, and multilingual text; 4K output up to 8.29 megapixels; and natural multi-image blending with consistent scale, perspective, and lighting.

What resolutions does Seedream 5.5 support?

Custom width and height up to 8.29 megapixels, with 4K UHD (3840 × 2160) at 16:9 among the supported sizes. API presets of 1K, 1.5K, and 2K are convenience shortcuts, not the full resolution limit.

How many layers can Seedream 5.5 output?

Up to 99 layers per image, split into background, element, and text layers. Text layers come back as a PNG plus the text itself, structured and readable, so downstream tools can edit the copy for localization or template reuse.

Will Seedream 5.5 be on Morphic?

The Seedream line is already on Morphic, so 5.5 is expected to arrive here when ByteDance publishes the API. You can use the current Seedream model now and switch to 5.5 in the same workspace.

What can I use while Seedream 5.5 is on the way?

The current Seedream image model is live on Morphic today, with grounded editing, layer decomposition, and native multilingual text. Start there and move to 5.5 when it ships.

More about Seedream 5.5

Seedream 5.5: features, specs, and what ByteDance has announced

Seedream 5.5 is ByteDance's next image model. Multi-round editing, transparent output, 99-layer separation, multilingual text, and 4K at up to 8.29 MP.

Other models

Gemini Omni Flash 1.2

Google DeepMind

Google's video model, now fifteen seconds long. Edit the whole take for 30% less a second.

ChatGPT Images 2.6

OpenAI

OpenAI's image model, released November 2026. Animated GIFs, SVG export, faster drafts.

SAM 3.1