Image

Seedream5.5 API

Explore Seedream5.5 API workflows for image generation, precise edits, multilingual text, and multi-image composition. Use this APIMart workspace to organize prompts and references, then review the API examples and parameter table below.

Image generation & editingMultilingual textPrecise local changes

Get API Key API Docs

Seedream5.5 API workspace

1K, 1.5K, 2K, aspect ratios, and custom pixels
Up to 10 references; one for transparency or layers
JPEG / PNG, prompt optimization, and output settings

Input

Model

Prompt

Describe the subject, composition, lighting, and any text to include.

Size

Resolution

Output Format

Prompt Mode

Background

Content Moderation

Watermark

Reference images

Up to 10 references, or one for transparency or layer separation. Local files: JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC, or HEIF, up to 30 MB each. Use public HTTP(S) URLs or complete image data URIs in JSON; local files remain browser previews.

Reference images: each side greater than 14 px, width-to-height ratio 1/16–16, and at most 36,000,000 total pixels. Transparent backgrounds require an input with an alpha channel.

Output

Your generated images will appear here

Create a clear visual brief with Seedream5.5 API

Prepare a Seedream5.5 API draft for a new scene or a focused edit. Define the subject, light, and composition, then separate requested changes from details to preserve.

Get API key

Seedream5.5 API model capabilities

Use these Seedream5.5 capabilities to shape your API prompts, references, and editing instructions.

Precise edits with Seedream5.5

Seedream5.5 supports focused changes to objects, expressions, and text. In your API prompt, identify the region to edit and the surrounding content to preserve.

Multi-round consistency with Seedream 5.5

Build each Seedream5.5 revision from the current input image. Include the previous result again when continuing an edit, and identify the subject, layout, and text to retain.

People, materials, and light in the Seedream 5.5 model

Seedream5.5 addresses facial expressions, proportions, materials, and scene lighting. Describe the intended expression, surface finish, and light direction instead of relying on a generic request for realism.

Multi-image composition with seedream-5.5

Give each reference a role in your Seedream5.5 API draft: subject, setting, or style. Specify relative scale, perspective, and lighting so the separate sources form one composition.

Dense and multilingual text with Seedream5.5 API

Provide exact wording, hierarchy, and placement in the API prompt. Seedream5.5 supports small text and multilingual layouts; review spelling and legibility at the intended display size.

Transparent assets with ByteDance Seedream5.5

Seedream5.5 covers transparent backgrounds and semitransparent assets. Specify an isolated element or a translucent material such as glass; these are different requirements for the background and subject.

Layer separation in Seedream5.5

Seedream5.5 distinguishes backgrounds, visual elements, and text for further editing. Name the parts to separate in your API prompt, including their alignment and any wording to revise.

From sketch to image with seedream-5.5

Add a sketch as a Seedream5.5 reference and describe the intended materials, style, and light. Identify which outlines and positions should remain in the finished image.

Creative workflows for Seedream5.5 API

Connect each Seedream5.5 API brief to a concrete design task and review checklist.

Campaign posters with Seedream5.5

Structure a Seedream5.5 API prompt around the headline, supporting copy, and main visual. Define reading order, brand space, and separate text blocks for each language.

Product displays with ByteDance Seedream5.5

Use Seedream5.5 references to preserve packaging, labels, and materials. Describe the new studio or lifestyle setting, including reflections and contact shadows around the product.

Stickers and icons with seedream-5.5

Define silhouettes, edge treatment, and background transparency in your Seedream5.5 API brief. Distinguish soft or semitransparent subject details from the empty background around them.

Infographics and slides with Seedream5.5 API

Supply Seedream5.5 with the actual facts, numbers, and labels. Use the API prompt to organize headings and diagrams into a readable hierarchy; check factual accuracy as well as layout.

Storyboard planning with the Seedream 5.5 model

Plan each Seedream5.5 shot with its action, camera position, and lighting. Include character features, clothing, recurring props, and the relevant reference image to guide continuity.

Local retouching and text changes with Seedream5.5

Pair old text with its replacement in the Seedream5.5 API prompt. Identify the edit location and preserve the surrounding layout, colors, and subjects.

Prepare a Seedream5.5 API creative brief

Review the prompt, references, and API fields together before copying your configuration.

Write your Seedream5.5 API prompt

Describe the scene or the specific edit in prompt. Quote exact image text, name the location of each change, and list details to preserve.

→

Prepare references for seedream-5.5

Assign each reference a role, then choose size, resolution, and output format. Transparency and layer separation use one reference; local files remain browser previews.

→

Review the Seedream5.5 API JSON

Review the Seedream5.5 API prompt, references, size, and output options in Form or JSON. Check the single-image requirements for transparent backgrounds and layer separation.

Get API key

Seedream5.5 API examples: text to image and image editing

The Seedream5.5 API examples use APIMart’s common asynchronous task format with prompts, references, size, resolution, and output options. Use the returned task ID for a separate result query.

1
Set up your API key
Save your APIMart API key as APIMART_API_KEY. The Seedream5.5 API examples use Authorization: Bearer <key> for both submission and task queries.
2
Build the Seedream5.5 API request
The example uses POST /v1/images/generations with model set to seedream-5.5, a prompt, size, resolution, and output_format. Add image_urls for editing. Read the task ID from data[0].task_id.
3
Read the API task result
Use GET /v1/tasks/{task_id}. For a completed task, read data.result.images[0].url[0]; for a failed task, read data.error.message.
See the APIMart documentation for APIMart authentication and general API usage.

POST /v1/images/generations

curl -X POST https://api.apimart.ai/v1/images/generations \
  -H "Authorization: Bearer $APIMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedream-5.5",
    "prompt": "An editorial product poster featuring a white ceramic cup on a pale wooden table, soft side lighting, headline \"MORNING RITUAL\" at the top, clear typography and generous whitespace.",
    "size": "16:9",
    "resolution": "1.5K",
    "output_format": "png"
  }'

import os
import time
import requests

API_BASE = "https://api.apimart.ai"
headers = {"Authorization": f"Bearer {os.environ['APIMART_API_KEY']}"}

# 1. Submit the generation task
payload = {
  "model": "seedream-5.5",
  "prompt": "An editorial product poster featuring a white ceramic cup on a pale wooden table, soft side lighting, headline \"MORNING RITUAL\" at the top, clear typography and generous whitespace.",
  "size": "16:9",
  "resolution": "1.5K",
  "output_format": "png"
}
resp = requests.post(f"{API_BASE}/v1/images/generations", headers=headers, json=payload, timeout=60)
resp.raise_for_status()
task_id = resp.json()["data"][0]["task_id"]

# 2. Poll the task until it finishes
for _ in range(200):
    task_resp = requests.get(f"{API_BASE}/v1/tasks/{task_id}", headers=headers, timeout=30)
    task_resp.raise_for_status()
    task = task_resp.json()["data"]
    if task["status"] == "completed":
        print(task["result"]["images"][0]["url"][0])
        break
    if task["status"] == "failed":
        raise RuntimeError(task.get("error", {}).get("message", "generation failed"))
    time.sleep(3)
else:
    raise TimeoutError(f"task {task_id} did not finish in time; continue GET polling with this task ID")

curl -X POST https://api.apimart.ai/v1/images/generations \
  -H "Authorization: Bearer $APIMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedream-5.5",
    "prompt": "In the reference image, replace the headline \"MORNING RITUAL\" with \"EVENING CALM\". Preserve the cup, table, lighting, text position, and all other elements.",
    "image_urls": [
      "https://example.com/reference.png"
    ],
    "size": "auto",
    "resolution": "2K",
    "output_format": "png"
  }'

TASK_ID="task_xxx"  # replace with the task_id returned by the POST request
curl "https://api.apimart.ai/v1/tasks/$TASK_ID" \
  -H "Authorization: Bearer $APIMART_API_KEY"

Seedream5.5 API parameters

Parameter	Type	Required	Default	Description
`model`	string	Required	Default: seedream-5.5	The model value in this Seedream5.5 API workspace and its examples: seedream-5.5.
`prompt`	string	Required	Default: —	Your Seedream5.5 instruction: describe a new image, or identify edits and details to preserve. Optional when layer_decomposition is true.
`size`	string	Optional	Default: 1:1	Workspace default: 1:1. Choose auto; 1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, 2:1, 1:2, 21:9; a tier (1K, 1.5K, 2K); or custom pixels such as 2048x1024. Custom sizes require 921,600–4,624,220 total pixels and a width-to-height ratio of 1/16–16.
`resolution`	string	Optional	Default: 1K	Seedream5.5 API draft resolution: 1K, 1.5K, or 2K. Applies when size is a ratio or auto. Omitted for tier or pixel sizes and for layer decomposition.
`image_urls`	array	Optional	Default: —	Up to 10 reference images as public HTTP(S) URLs or complete base64 image data URIs. JPEG, PNG, WebP, BMP, TIFF, GIF, HEIC, and HEIF; up to 30 MB each. Use an aspect ratio of 1/16–16, each side greater than 14 px, and at most 36,000,000 total pixels. Transparency or layer decomposition uses exactly one reference. Local blob URLs are browser previews.
`output_format`	string	Optional	Default: jpeg	Output format: jpeg or png. Selecting a transparent background in Form automatically selects PNG.
`background`	string	Optional	Default: opaque	opaque or transparent. A transparent background requires exactly one input image with an alpha channel and PNG output.
`layer_decomposition`	boolean	Optional	Default: false	Split one PNG or JPEG reference into a base image and transparent layers. In this mode, size accepts only auto, 1K, 1.5K, or 2K; prompt is optional and resolution is omitted.
`optimize_prompt_options`	object	Optional	Default: Omitted	Optional prompt optimization: {"mode":"standard"}. The workspace defaults to Off, which omits this field.
`watermark`	boolean	Optional	Default: false	Whether to add an AI-generated watermark to the output image.
`nsfw_check`	boolean	Optional	Default: false	Whether to moderate the prompt and input images before processing the API task.
`n`	integer	Optional	Default: 1	Normal generation produces one image. JSON accepts only n: 1; the workspace omits it from its normalized configuration.

Seedream5.5 API FAQ

Seedream5.5 prompts, editing decisions, API keys, request fields, and JSON configuration.

What does Seedream5.5 focus on?

Seedream5.5 is ByteDance’s image-generation and editing model. Its capabilities cover local changes, consistency through revisions, people and materials, multi-image composition, multilingual text, transparent assets, and layer separation.

How should I write a seedream-5.5 prompt?

Start a Seedream5.5 API prompt with the subject and scene, then add composition, materials, and light. For editing, identify the change and what must remain. Use concrete positions rather than vague instructions such as “improve it.”

How do I request a precise local edit with Seedream5.5 API?

Identify the object and its location in the API prompt: for example, change the blue cup on the left to white ceramic while preserving the table and framing. Review surrounding details as well as the edited region.

Does the Seedream 5.5 model automatically remember earlier images?

No automatic image memory should be assumed. Supply the image to edit for each Seedream5.5 revision, including a previous result or reference when needed. In the API example, list these inputs in image_urls and describe their roles.

How do I combine references with ByteDance Seedream5.5?

Assign reference roles in your Seedream5.5 API prompt: subject from image A, setting from B, and color treatment from C. Specify relative size, viewpoint, placement, and light direction so the inputs fit together.

What should I specify when editing a person with Seedream5.5?

Give Seedream5.5 the intended expression or pose and list facial features, clothing, proportions, and framing to retain. If only the expression should change, state that limit explicitly and inspect the rest of the person afterward.

How can I prepare small or dense text for Seedream5.5 API?

Put the complete wording in quotation marks in the API prompt. Separate headlines, labels, and body copy into positioned blocks. Check Seedream5.5 text at its intended display size, especially digits, punctuation, and repeated labels.

How should I organize multilingual text in seedream-5.5?

Give Seedream5.5 the exact copy for each language and identify related text blocks. Specify placement and reading order in the API prompt, then review spelling and layout in every language.

How do I replace wording while preserving a Seedream5.5 layout?

Use an explicit replacement in your Seedream5.5 API prompt, such as “SUMMER SALE” to “AUTUMN SALE.” Name the text location and preserve surrounding graphics, alignment, and other wording. Check that the old text is fully removed.

Is a transparent background the same as a translucent object in Seedream5.5?

They describe different requirements. In this Seedream5.5 API draft, background: transparent requires exactly one input image with an alpha channel and output_format: png. A glass object can instead appear against an opaque, fully visible background.

What details help describe glass and other materials to the Seedream 5.5 model?

Specify whether the surface is clear, frosted, glossy, matte, or metallic. Describe the light direction, reflections, and what should be visible through a translucent object. Mention any material details from the reference that must remain.

What does layer separation mean for Seedream5.5?

Seedream5.5 separates a reference into a base image and transparent layers. Select Split Layers or set layer_decomposition: true, supply one PNG or JPEG reference, and choose size auto, 1K, 1.5K, or 2K. The API draft omits resolution in this mode; prompt is optional.

Where do I get an API key for the Seedream5.5 API examples?

Use Get API key to open APIMart’s key management page. Save your key in the server environment as APIMART_API_KEY and use it in the Authorization header shown in the examples. Keep API keys out of prompts, reference URLs, and public code.

Which fields do the Seedream5.5 API examples use?

The Seedream5.5 API examples use model, prompt, image_urls, size, resolution, and output_format. Additional options cover background, layer_decomposition, optimize_prompt_options, watermark, and nsfw_check. Normal generation uses one output image; n accepts only 1. The parameter table lists the defaults and field dependencies.

Are seedream-5-5-api and Seedream5.5 different models?

These are name variants used on this page for ByteDance Seedream5.5, also written Seedream 5.5. A search phrase such as seedream-5-5-api is not a request alias. The model value in this workspace and its API examples is seedream-5.5.

How do Form and JSON work in the Seedream5.5 API workspace?

Both views edit the same local configuration, including prompts, references, size, and output options. Invalid JSON, unsupported fields, or incompatible parameter combinations show an error without replacing the last valid draft. Seedream5.5 API image inputs use reachable URLs or complete image data URIs; local blob URLs remain browser previews.

Related Models

Explore more models in the same category.

View All Models →

Seedream5.5 API FAQ

Seedream5.5 prompts, editing decisions, API keys, request fields, and JSON configuration.

What does Seedream5.5 focus on?

How should I write a seedream-5.5 prompt?

How do I request a precise local edit with Seedream5.5 API?

Does the Seedream 5.5 model automatically remember earlier images?

How do I combine references with ByteDance Seedream5.5?

What should I specify when editing a person with Seedream5.5?

How can I prepare small or dense text for Seedream5.5 API?

How should I organize multilingual text in seedream-5.5?

Give Seedream5.5 the exact copy for each language and identify related text blocks. Specify placement and reading order in the API prompt, then review spelling and layout in every language.

Seedream5.5 API

Create a clear visual brief with Seedream5.5 API

Seedream5.5 API model capabilities

Creative workflows for Seedream5.5 API

Prepare a Seedream5.5 API creative brief

Write your Seedream5.5 API prompt

Prepare references for seedream-5.5

Review the Seedream5.5 API JSON

Seedream5.5 API examples: text to image and image editing

Set up your API key

Build the Seedream5.5 API request

Read the API task result

Seedream5.5 API parameters

Seedream5.5 API FAQ

What does Seedream5.5 focus on?

How should I write a seedream-5.5 prompt?

How do I request a precise local edit with Seedream5.5 API?

Does the Seedream 5.5 model automatically remember earlier images?

How do I combine references with ByteDance Seedream5.5?

What should I specify when editing a person with Seedream5.5?

How can I prepare small or dense text for Seedream5.5 API?

How should I organize multilingual text in seedream-5.5?

How do I replace wording while preserving a Seedream5.5 layout?

Is a transparent background the same as a translucent object in Seedream5.5?

What details help describe glass and other materials to the Seedream 5.5 model?

What does layer separation mean for Seedream5.5?

Where do I get an API key for the Seedream5.5 API examples?

Which fields do the Seedream5.5 API examples use?

Are seedream-5-5-api and Seedream5.5 different models?

How do Form and JSON work in the Seedream5.5 API workspace?

Related Models

Seedream5.5 API model capabilities

Creative workflows for Seedream5.5 API

Prepare a Seedream5.5 API creative brief

Write your Seedream5.5 API prompt

Prepare references for seedream-5.5

Review the Seedream5.5 API JSON

Seedream5.5 API FAQ

What does Seedream5.5 focus on?

How should I write a seedream-5.5 prompt?

How do I request a precise local edit with Seedream5.5 API?

Does the Seedream 5.5 model automatically remember earlier images?

How do I combine references with ByteDance Seedream5.5?

What should I specify when editing a person with Seedream5.5?

How can I prepare small or dense text for Seedream5.5 API?

How should I organize multilingual text in seedream-5.5?

How do I replace wording while preserving a Seedream5.5 layout?

Is a transparent background the same as a translucent object in Seedream5.5?

What details help describe glass and other materials to the Seedream 5.5 model?

What does layer separation mean for Seedream5.5?

Where do I get an API key for the Seedream5.5 API examples?

Which fields do the Seedream5.5 API examples use?

Are seedream-5-5-api and Seedream5.5 different models?

How do Form and JSON work in the Seedream5.5 API workspace?

Related Models