An all-round upgrade through scaling
By scaling up the model overall, it improves consistency and fidelity across generation and editing compared to Seedream 4.
Image by ByteDance
Seedream 4.5 API
ByteDance's refined image generation and editing model (Dec 2025), an all-round upgrade of Seedream 4 with best-in-class text rendering, stricter reference preservation, and designer-level typography.
Integrate Seedream 4.5 with a single API call — one key, one unified endpoint, and shared billing across every model on Apiframe.
model: "seedream-4.5" 
What's special about Seedream 4.5
Best-in-class text rendering
It spells accurately, supports varied fonts, handles curved and rotated text, places multiple text elements cleanly, and keeps small text readable.
Strict reference preservation
In multi-image editing it identifies the main subjects accurately and preserves facial features, lighting, and color tone from the reference images.
Strong multi-reference support
It processes up to 14 reference images at once, holding consistent subjects, lighting, and style across complex compositions.
Designer-level composition
It delivers professional layout and typography at high resolution up to 4K, suitable for posters, brand visuals, and print.
Consistent sets and flexible styles
It generates multiple related images with consistent characters and themes, and transforms across a wide range of artistic styles.
Made with Seedream 4.5
A few outputs generated through the Seedream 4.5 API on Apiframe.
A modern e-commerce hot-sale image with the title 'Today's Hot Sale', a product centered, price '$9.99' below, orange-to-purple gradient.
A minimalist art exhibition poster, off-white background, a black line-drawing portrait slightly right of center.
A coffee shop menu board, chalkboard style, title 'DAILY SPECIALS' with Espresso $3, Latte $4, Cappuccino $4.50.
Using these reference photos, keep the model's face and lighting identical and change the dress to flowing red silk.
A European vintage wedding invitation, dark green base, gold-foil typography with names, date, and venue.
A product line-up of the same sneaker in six colorways, consistent angle, lighting, and style.
Overview
- Endpoint
- POST /v2/images/generate
- Model ID
- seedream-4.5
- Params key
- seedreamParams
- Modality
- Image
- Provider
- ByteDance
- Avg. completion
- ~15s
Capabilities
Aspect ratios1:1, 3:4, 4:3, 9:16, 16:9
Image inputSupported
Avg. time~15s
Quick start
Send a single POST /v2/images/generate request with your API key to
generate with Seedream 4.5. The call returns a jobId you can poll or
receive via webhook.
curl -X POST https://api.apiframe.ai/v2/images/generate \
-H "X-API-Key: afk_your_api_key_here" \
-H "Content-Type: application/json" \
-d '{
"prompt": "a sleek silver sports car on a coastal highway at sunset, hyper-realistic",
"model": "seedream-4.5",
"seedreamParams": {
"image_input": "https://example.com/input.jpg",
"output_format": "jpg",
"guidance_scale": 3,
"use_pre_llm": false
}
}'import requests
response = requests.post(
"https://api.apiframe.ai/v2/images/generate",
headers={
"X-API-Key": "afk_your_api_key_here",
"Content-Type": "application/json",
},
json={
"prompt": "a sleek silver sports car on a coastal highway at sunset, hyper-realistic",
"model": "seedream-4.5",
"seedreamParams": {
"image_input": "https://example.com/input.jpg",
"output_format": "jpg",
"guidance_scale": 3,
"use_pre_llm": False
}
},
)
print(response.json()) # { "jobId": "...", "status": "QUEUED" }const response = await fetch("https://api.apiframe.ai/v2/images/generate", {
method: "POST",
headers: {
"X-API-Key": "afk_your_api_key_here",
"Content-Type": "application/json",
},
body: JSON.stringify({
"prompt": "a sleek silver sports car on a coastal highway at sunset, hyper-realistic",
"model": "seedream-4.5",
"seedreamParams": {
"image_input": "https://example.com/input.jpg",
"output_format": "jpg",
"guidance_scale": 3,
"use_pre_llm": false
}
}),
});
const { jobId } = await response.json();
console.log(jobId);Response & job lifecycle
Generation is asynchronous. A successful submission returns 202 Accepted with a jobId. Poll GET /v2/jobs/{id} (or supply a webhook_url) until the status is COMPLETED; the result field then holds the output URL(s).
1. Submission response (202)
{
"jobId": "b2c3d4e5-f6a7-8901-bcde-f23456789012",
"status": "QUEUED"
}2. Poll for the result
curl https://api.apiframe.ai/v2/jobs/JOB_ID \
-H "X-API-Key: afk_your_api_key_here"import requests, time
while True:
job = requests.get(
"https://api.apiframe.ai/v2/jobs/JOB_ID",
headers={"X-API-Key": "afk_your_api_key_here"},
).json()
if job["status"] in ("COMPLETED", "FAILED"):
break
time.sleep(2)
print(job["result"])let job;
do {
await new Promise((r) => setTimeout(r, 2000));
job = await fetch("https://api.apiframe.ai/v2/jobs/JOB_ID", {
headers: { "X-API-Key": "afk_your_api_key_here" },
}).then((r) => r.json());
} while (job.status !== "COMPLETED" && job.status !== "FAILED");
console.log(job.result);Input schema
Request parameters accepted by the Seedream 4.5 endpoint. Model-specific options are nested under the params object shown below.
| Parameter | Type | Required | Default | Allowed / range | Description |
|---|---|---|---|---|---|
| prompt | string | required | — | — | Text description of what to generate. |
| model | string | required | "seedream-4.5" | "seedream-4.5" | The model identifier for this endpoint. |
| seedreamParams.image_input | string (URL) | optional | — | — | Reference image (URL) |
| seedreamParams.output_format | string | optional | "jpg" | "jpg", "png" | Output format |
| seedreamParams.guidance_scale | number | optional | 3 | min 1, max 10, step 0.5 | Guidance |
| seedreamParams.use_pre_llm | boolean | optional | false | — | Pre-process the prompt with an LLM. |
| seedreamParams.enhance_prompt | boolean | optional | false | — | Enhance prompt |
| seedreamParams.seed | number | optional | — | step 1 | Reuse a number to reproduce the same result. |
Frequently Asked Questions
Common questions about the Seedream 4.5 API.
What is Seedream 4.5?
ByteDance's latest image generation and editing model, released in December 2025 as an all-round upgrade of Seedream 4.
How is it different from Seedream 4?
It scales up the model for better multi-image subject identification, stricter reference preservation, and stronger typography and dense-text rendering.
What is it best at?
Text rendering and designer-level typography, across posters, brand visuals, and e-commerce imagery.
How many reference images can it use?
Up to 14 at once, while keeping subjects, lighting, and style consistent.
How much does it cost?
On Apiframe, 8 credits ($0.08) per image.
Where can you access it?
Through Apiframe, as well as ByteDance's platforms and major hosting partners.
Still have questions?
Start building with the Seedream 4.5 API
Get your API key and integrate Seedream 4.5 in minutes — Pay-as-you-go.
Free credits to start
One API for every model
Webhooks, SDKs & idempotency
No provider account required
Questions? Join our Discord or contact sales.