Seedream 4.5

Overview

Seedream 4.5 is a ByteDance image model for text-to-image generation and reference-guided image generation. It supports 2K and 4K output tiers, up to 10 images per request, and optional sequential image generation for connected image sets. Use this page when you are ready to call the API after trying the model in the APIXO playground.

Capability	Value
Model ID	`seedream-4-5`
Modes	`text-to-image`, `image-to-image`
Prompt length	1-5000 characters
Reference images	1-14 URLs for `image-to-image`
Images per task	`max_images` is required, 1-10
Resolution tiers	`2K`, `4K`
Aspect ratios	`1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `9:16`, `16:9`, `21:9`
Sequential generation	`disabled`, `auto`

Endpoint and authentication

Base URL:

https://api.apixo.ai/api/v1

Method	Endpoint	Purpose
`POST`	`/generateTask/seedream-4-5`	Submit a generation task
`GET`	`/statusTask/seedream-4-5?taskId={taskId}`	Poll task status and retrieve results

All requests require your APIXO API key:

Authorization: Bearer YOUR_API_KEY

Submit requests also require:

Content-Type: application/json

Copy-paste async quickstart

This minimal request submits a text-to-image task and returns a taskId.

curl -X POST "https://api.apixo.ai/api/v1/generateTask/seedream-4-5" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "request_type": "async",
    "input": {
      "mode": "text-to-image",
      "prompt": "a cozy cyberpunk cafe with warm neon lights, ultra detailed",
      "max_images": 2,
      "resolution": "2K",
      "aspect_ratio": "16:9",
      "sequential_image_generation": "disabled"
    }
  }'

Successful response:

{
  "code": 200,
  "message": "success",
  "data": {
    "taskId": "task_12345678"
  }
}

Save the taskId; you need it to poll for the final result.

Poll for result

curl -X GET "https://api.apixo.ai/api/v1/statusTask/seedream-4-5?taskId=task_12345678" \
  -H "Authorization: Bearer YOUR_API_KEY"

Processing response:

{
  "code": 200,
  "message": "success",
  "data": {
    "taskId": "task_12345678",
    "state": "processing",
    "createTime": 1767965610929
  }
}

Success response:

{
  "code": 200,
  "message": "success",
  "data": {
    "taskId": "task_12345678",
    "state": "success",
    "resultJson": "{\"resultUrls\":[\"https://r2.apixo.ai/img1.png\",\"https://r2.apixo.ai/img2.png\"]}",
    "createTime": 1767965610929,
    "completeTime": 1767965652317,
    "costTime": 41388
  }
}

Failed response:

{
  "code": 200,
  "message": "success",
  "data": {
    "taskId": "task_12345678",
    "state": "failed",
    "failCode": "SensitiveContentDetected",
    "failMsg": "The input or output was flagged as sensitive. Please try again with different inputs.",
    "createTime": 1767965610929,
    "completeTime": 1767965620132
  }
}

Parse resultJson after state becomes success:

const payload = JSON.parse(data.resultJson);
const imageUrls = payload.resultUrls;

Request body

Text-to-image

{
  "request_type": "async",
  "input": {
    "mode": "text-to-image",
    "prompt": "a fox in watercolor style",
    "max_images": 1,
    "resolution": "2K",
    "aspect_ratio": "1:1",
    "sequential_image_generation": "disabled"
  }
}

Image-to-image

{
  "request_type": "async",
  "input": {
    "mode": "image-to-image",
    "prompt": "turn this scene into a warm sunrise illustration",
    "image_urls": [
      "https://example.com/reference-1.png",
      "https://example.com/reference-2.jpg"
    ],
    "max_images": 3,
    "resolution": "4K",
    "aspect_ratio": "3:4",
    "sequential_image_generation": "auto"
  }
}

Multi-image sequence

{
  "request_type": "async",
  "input": {
    "mode": "text-to-image",
    "prompt": "four connected storyboard frames of a small robot exploring a greenhouse",
    "max_images": 4,
    "resolution": "2K",
    "aspect_ratio": "16:9",
    "sequential_image_generation": "auto"
  }
}

Parameters

request_type

string

default:"async"

Result delivery mode. Use async for polling with statusTask, or callback for webhook delivery.

callback_url

string

Required when request_type is callback. Must be a public HTTPS URL that can receive the final task payload. See Webhooks.

input

object

required

Seedream 4.5 input parameters.

Show properties

mode

string

required

Generation mode. Supported values: text-to-image, image-to-image.

prompt

string

required

Text prompt describing the desired image or edit. Supports 1-5000 characters after trimming whitespace.

max_images

integer

required

Number of images to request. Supported range: 1-10. This field is required; the backend does not apply a default.

image_urls

string[]

Reference image URLs. Required for image-to-image. Supports 1-14 URLs. URLs should be public, directly accessible image files.

resolution

string

default:"2K"

Output resolution tier. Supported values: 2K, 4K.

aspect_ratio

string

Optional output aspect ratio. Supported values: 1:1, 2:3, 3:2, 4:3, 3:4, 9:16, 16:9, 21:9. If omitted, APIXO sends only the selected resolution tier upstream instead of forcing an aspect ratio.

sequential_image_generation

string

default:"disabled"

Sequential image mode. Supported values: disabled, auto. Use auto when multiple images should be generated as a related sequence.

When aspect_ratio is provided, APIXO maps resolution and aspect_ratio to an exact upstream size:

Aspect ratio	`2K` size	`4K` size
`1:1`	`2048x2048`	`4096x4096`
`2:3`	`1664x2496`	`3328x4992`
`3:2`	`2496x1664`	`4992x3328`
`3:4`	`1728x2304`	`3520x4704`
`4:3`	`2304x1728`	`4704x3520`
`9:16`	`1600x2848`	`3040x5504`
`16:9`	`2848x1600`	`5504x3040`
`21:9`	`3136x1344`	`6240x2656`

Response format

Submit task response

POST /generateTask/seedream-4-5 returns a task ID when the task is accepted:

code

integer

API status code. 200 means the task was accepted.

message

string

Human-readable status message.

data.taskId

string

Unique task identifier used with the status endpoint.

Status response fields

taskId

string

Unique task identifier.

state

string

Current task state: pending, processing, success, or failed.

resultJson

string

JSON string containing the generated image URLs. Present when state is success.

failCode

string

Machine-readable failure code. Present when state is failed.

failMsg

string

Human-readable failure message. Present when state is failed.

createTime

integer

Task creation timestamp in Unix milliseconds.

completeTime

integer

Task completion timestamp in Unix milliseconds. Present after completion.

costTime

integer

Processing duration in milliseconds. Present after successful completion.

If at least one image succeeds, the task is returned as success and resultJson.resultUrls contains only the successful images. Any requested images that did not complete are not included in resultUrls.

Webhook callback mode

Use callback mode when your backend should receive the final result automatically instead of polling.

curl -X POST "https://api.apixo.ai/api/v1/generateTask/seedream-4-5" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "request_type": "callback",
    "callback_url": "https://your-server.com/webhooks/apixo",
    "input": {
      "mode": "text-to-image",
      "prompt": "a clean product photo of a ceramic teapot on soft stone",
      "max_images": 1,
      "resolution": "2K",
      "aspect_ratio": "1:1",
      "sequential_image_generation": "disabled"
    }
  }'

Successful callback payload:

{
  "code": 200,
  "message": "success",
  "data": {
    "taskId": "task_12345678",
    "state": "success",
    "resultJson": "{\"resultUrls\":[\"https://r2.apixo.ai/img1.png\"]}",
    "createTime": 1767965610929,
    "completeTime": 1767965652317,
    "costTime": 41388
  }
}

Failed callback payloads use code: 500, message: "failed", and include failCode / failMsg in data. See Webhooks for delivery requirements and retry behavior.

Billing

Seedream 4.5 is billed per successfully generated image.

Unit	APIXO price
Generated image	`$0.035 / image`

The backend pre-charges max_images * unit price when the task is accepted. If fewer images complete successfully, the unused portion is refunded automatically. For current route and market comparison pricing, see Pricing.

Latency and polling

Actual latency may vary by prompt complexity, number of requested images, reference image count, and current provider queue load.

Resolution	Typical generation time	Recommended first poll	Poll interval
`2K`	25s-35s	20s after task creation	3s-5s
`4K`	40s-55s	30s after task creation	5s

For high-concurrency production workloads, use callback mode to avoid frequent polling.

Seedream 4.5 streaming requests have an upstream processing timeout of about 180 seconds. If a task times out, retry with a simpler prompt, fewer reference images, or fewer requested images.

Result URLs are valid for 15 days. Download and store important outputs promptly. Rate limits and concurrency can vary by account, API key, and route. If you receive 429, slow down requests and retry with backoff. For account-level details, see System APIs.

Errors and troubleshooting

HTTP errors

Code	Meaning	What to do
`400`	Invalid request body, mode, parameter type, parameter value, or prompt length	Fix the request before retrying
`401`	Missing or invalid API key	Check the `Authorization` header
`402`	Insufficient balance or quota	Add balance or switch account/key
`403`	Key or route cannot access the model, or content is blocked before task creation	Check permissions and input content
`404`	Model route or task ID was not found	Check the endpoint path and `taskId`
`429`	Rate limit or concurrency limit reached	Retry with exponential backoff
`500`	Server error or streaming finalization error	Retry with backoff
`502`	Upstream provider error	Retry with backoff
`504`	Upstream timeout	Retry, reduce input complexity, or use callback mode

Task failure codes

Fail code	Meaning	What to do
`SensitiveContentDetected`	Input or output was flagged as sensitive	Change the prompt or reference images
`InputOutputSensitiveContentDetected`	Input or output failed safety checks	Use different prompt or image inputs
`PromptInvalid`	Prompt was invalid or rejected by the provider	Rewrite the prompt and retry
`BadRequest`	Required parameters were missing or malformed upstream	Check `mode`, `prompt`, `max_images`, `resolution`, `aspect_ratio`, and `image_urls`
`InvalidImageSize`	A reference image size is unsupported	Use a different image size
`ImageFormatIncorrect`	A reference image format could not be accepted	Use a direct JPEG, PNG, or WebP image URL
`Upload error`	A reference image is too large	Compress the image and retry
`RateLimitExceeded`	Provider-side rate limit reached	Retry later with backoff
`Timeout`	Provider processing timed out	Retry with simpler inputs or fewer images
`StreamError`	Streaming request failed after retries	Retry later
`FinalizeError`	Result finalization failed	Retry or contact support with the `taskId`
`Validation error`	Upstream rejected the request as invalid	Check all parameters and image URLs
`UpstreamError`	Provider returned an unmapped failure	Retry or contact support with the `taskId`

Common fixes

Missing mode: Include input.mode; routing requires text-to-image or image-to-image.
Missing max_images: Include input.max_images; it is required even when you only want one image.
Image-to-image errors: Include at least one public, direct URL in image_urls; the maximum is 14 URLs.
Unexpected aspect ratio: Omit aspect_ratio only if you want the provider to use the resolution tier without a forced ratio.
Partial output: If only some images complete, the task can still be success; parse resultUrls and use the returned URLs only. See Error Codes for the full error reference.

Getting Started

Image Models

Video Models

Audio Models

Seedream 4.5

Overview

Endpoint and authentication

Copy-paste async quickstart

Poll for result

Request body

Text-to-image

Image-to-image

Multi-image sequence

Parameters

Response format

Submit task response

Status response fields

Webhook callback mode

Billing

Latency and polling

Errors and troubleshooting

HTTP errors

Task failure codes

Common fixes

Getting Started

Image Models

Video Models

Audio Models

Documentation Index

​Overview

​Endpoint and authentication

​Copy-paste async quickstart

​Poll for result

​Request body

​Text-to-image

​Image-to-image

​Multi-image sequence

​Parameters

​Response format

​Submit task response

​Status response fields

​Webhook callback mode

​Billing

​Latency and polling

​Errors and troubleshooting

​HTTP errors

​Task failure codes

​Common fixes

​Related links

Overview

Endpoint and authentication

Copy-paste async quickstart

Poll for result

Request body

Text-to-image

Image-to-image

Multi-image sequence

Parameters

Response format

Submit task response

Status response fields

Webhook callback mode

Billing

Latency and polling

Errors and troubleshooting

HTTP errors

Task failure codes

Common fixes

Related links