Seedream 5.0 Lite API: Hurtigste endpoint til billedgenerering

REST API, Bearer-godkendelse, asynkron generering, webhooks og native batching op til 50 billeder pr. anmodning. Alt, der er tilgængeligt i Arteza-webinterfacet, findes også i API'et. Her er den fulde reference med fungerende Python- og JavaScript-eksempler, du kan kopiere direkte til produktion i dag.

API-oversigt

Seedream 5.0 Lite API giver programmatisk adgang til billedgenereringspipelinen på Arteza-platformen. Alt i webinterfacet, herunder tekst-til-billede, deep thinking, stiltransfer og tekstrendering, findes i REST API'et.

API'et følger REST-konventioner med JSON i både anmodnings- og svarobjekter. Generering er asynkron: send en anmodning, modtag et opgave-ID, og poll derefter eller modtag en webhook, når processen er færdig.

Fuld funktionsoversigt i vores komplet guide.

Basis-URL

https://api.arteza.ai/v1

Vigtige egenskaber

Godkendelse

Bearer token-godkendelse. Generér din API-nøgle fra Arteza-dashboardet under Indstillinger > API-nøgler.

Hent din API-nøgle

1. Tilmeld dig eller log ind på din Arteza-konto
2. Gå til Indstillinger > API-nøgler
3. Klik på Generér ny nøgle
4. Kopiér og opbevar din nøgle sikkert (vises kun én gang)

Godkendelsesheader

Authorization: Bearer YOUR_API_KEY

Test af godkendelse

curl -H "Authorization: Bearer YOUR_API_KEY" \
     https://api.arteza.ai/v1/account/credits

Svar:

{
  "credits_remaining": 1050,
  "tier": "starter"
}

Endpoint til enkelt billede

POST /v1/images/generate

Minimal anmodning

{
  "model": "seedream-5.0-lite",
  "prompt": "A serene mountain landscape at sunrise"
}

Fuld anmodning

{
  "model": "seedream-5.0-lite",
  "prompt": "Professional YouTube thumbnail with text 'TOP 10 TIPS' in bold red font, excited person on left, bright blue background",
  "aspect_ratio": "16:9",
  "deep_thinking": true,
  "style": "photorealistic",
  "webhook_url": "https://your-app.com/webhook/image-complete",
  "metadata": {
    "project": "youtube-thumbnails",
    "batch_id": "thumb-2026-04"
  }
}

Svar

{
  "task_id": "img_abc123def456",
  "status": "processing",
  "model": "seedream-5.0-lite",
  "credits_charged": 7,
  "credits_remaining": 1043,
  "estimated_time_seconds": 10
}

Anmodningsparametre

Billedformater

Stilpresetter

Svarformat

Polling for resultater

GET /v1/images/{task_id}

Behandler:

{
  "task_id": "img_abc123def456",
  "status": "processing",
  "progress": 0.65,
  "estimated_time_remaining": 5
}

Færdig:

{
  "task_id": "img_abc123def456",
  "status": "completed",
  "image_url": "https://cdn.arteza.ai/generated/img_abc123def456.png",
  "image_url_webp": "https://cdn.arteza.ai/generated/img_abc123def456.webp",
  "width": 1024,
  "height": 1024,
  "model": "seedream-5.0-lite",
  "deep_thinking_used": true,
  "credits_charged": 7,
  "metadata": {
    "project": "youtube-thumbnails",
    "batch_id": "thumb-2026-04"
  },
  "created_at": "2026-04-10T14:30:00Z"
}

Udløb af billed-URL

Genererede billed-URL'er er gyldige i 24 timer. Download og gem billeder i din egen lagring inden for dette tidsvindue.

Vil du have tekst, der ser sådan ud? Prøv Seedream 5.0 Lite gratis →

Python-eksempel

import requests
import time

API_KEY = "your_api_key_here"
BASE_URL = "https://api.arteza.ai/v1"

headers = {
    "Authorization": f"Bearer {API_KEY}",
    "Content-Type": "application/json"
}

# Send genereringsanmodning
response = requests.post(
    f"{BASE_URL}/images/generate",
    headers=headers,
    json={
        "model": "seedream-5.0-lite",
        "prompt": "A futuristic city skyline at sunset with flying cars",
        "aspect_ratio": "16:9",
        "deep_thinking": True
    }
)

task = response.json()
task_id = task["task_id"]
print(f"Opgave sendt: {task_id}")
print(f"Credits trukket: {task['credits_charged']}")
print(f"Resterende credits: {task['credits_remaining']}")

# Poll for færdiggørelse
while True:
    result = requests.get(
        f"{BASE_URL}/images/{task_id}",
        headers=headers
    ).json()

    if result["status"] == "completed":
        print(f"Billede klar: {result['image_url']}")
        break
    elif result["status"] == "failed":
        print(f"Generering mislykkedes: {result.get('error', 'Ukendt fejl')}")
        break

    time.sleep(2)

JavaScript-eksempel

const API_KEY = 'your_api_key_here';
const BASE_URL = 'https://api.arteza.ai/v1';

async function generateImage(prompt, options = {}) {
  const response = await fetch(`${BASE_URL}/images/generate`, {
    method: 'POST',
    headers: {
      'Authorization': `Bearer ${API_KEY}`,
      'Content-Type': 'application/json'
    },
    body: JSON.stringify({
      model: 'seedream-5.0-lite',
      prompt,
      aspect_ratio: options.aspectRatio || '1:1',
      deep_thinking: options.deepThinking || false,
      ...options
    })
  });

  const task = await response.json();
  console.log(`Opgave sendt: ${task.task_id}`);

  while (true) {
    const result = await fetch(
      `${BASE_URL}/images/${task.task_id}`,
      { headers: { 'Authorization': `Bearer ${API_KEY}` } }
    ).then(r => r.json());

    if (result.status === 'completed') {
      return result;
    }
    if (result.status === 'failed') {
      throw new Error(result.error || 'Generering mislykkedes');
    }

    await new Promise(resolve => setTimeout(resolve, 2000));
  }
}

// Brug
const result = await generateImage(
  'A cozy coffee shop interior with warm lighting',
  { aspectRatio: '16:9', deepThinking: true }
);
console.log(`Billed-URL: ${result.image_url}`);

Download af genererede billeder

import requests

def download_image(image_url, filename):
    response = requests.get(image_url)
    with open(filename, 'wb') as f:
        f.write(response.content)
    print(f"Gemt: {filename}")

download_image(result['image_url'], 'output/my_image.png')

Batch-genereringsendpoint

Til flere billeder i én anmodning:

POST /v1/images/batch

Batch-anmodning

{
  "generations": [
    {
      "prompt": "Minimalist logo design, blue circle with lightning bolt",
      "model": "seedream-5.0-lite",
      "aspect_ratio": "1:1",
      "deep_thinking": true
    },
    {
      "prompt": "Professional headshot background, soft gradient",
      "model": "seedream-5.0-lite",
      "aspect_ratio": "1:1",
      "deep_thinking": false
    },
    {
      "prompt": "YouTube thumbnail with text 'MUST WATCH' in red",
      "model": "seedream-5.0-lite",
      "aspect_ratio": "16:9",
      "deep_thinking": true
    }
  ],
  "webhook_url": "https://your-app.com/webhook/batch-complete"
}

Batch-svar

{
  "batch_id": "batch_xyz789",
  "status": "processing",
  "total_generations": 3,
  "total_credits_charged": 21,
  "credits_remaining": 1029
}

Poll af batch-status

GET /v1/images/batch/{batch_id}

{
  "batch_id": "batch_xyz789",
  "status": "completed",
  "results": [
    {
      "index": 0,
      "status": "completed",
      "task_id": "img_001",
      "image_url": "https://cdn.arteza.ai/generated/img_001.png"
    },
    {
      "index": 1,
      "status": "completed",
      "task_id": "img_002",
      "image_url": "https://cdn.arteza.ai/generated/img_002.png"
    },
    {
      "index": 2,
      "status": "completed",
      "task_id": "img_003",
      "image_url": "https://cdn.arteza.ai/generated/img_003.png"
    }
  ]
}

Batch-grænser

For batch-arbejdsgange, se vores guide til massegeneration.

Webhook-integration

Webhooks eliminerer polling. Når genereringen er færdig, sender API'et en POST til din URL.

Webhook-payload

{
  "event": "image.completed",
  "task_id": "img_abc123def456",
  "batch_id": "batch_xyz789",
  "status": "completed",
  "image_url": "https://cdn.arteza.ai/generated/img_abc123def456.png",
  "model": "seedream-5.0-lite",
  "credits_charged": 7,
  "metadata": {
    "project": "youtube-thumbnails"
  },
  "created_at": "2026-04-10T14:30:00Z"
}

Handler-eksempel (Python/Flask)

from flask import Flask, request, jsonify
import requests

app = Flask(__name__)

@app.route('/webhook/image-complete', methods=['POST'])
def handle_image_webhook():
    data = request.json

    if data['event'] == 'image.completed':
        task_id = data['task_id']
        image_url = data['image_url']
        metadata = data.get('metadata', {})

        # Download billede
        img_response = requests.get(image_url)
        filename = f"images/{metadata.get('project', 'default')}/{task_id}.png"
        with open(filename, 'wb') as f:
            f.write(img_response.content)

        # Opdatér din database
        update_generation_record(task_id, filename)

        print(f"Billede gemt: {filename}")

    elif data['event'] == 'image.failed':
        print(f"Generering mislykkedes: {data.get('error')}")

    return jsonify({'status': 'ok'}), 200

Webhook-sikkerhed

Bekræft webhook-autenticitet via X-Seedance-Signature-headeren:

import hmac
import hashlib

def verify_webhook(payload, signature, secret):
    expected = hmac.new(
        secret.encode(),
        payload.encode(),
        hashlib.sha256
    ).hexdigest()
    return hmac.compare_digest(expected, signature)

Deep thinking-tilstand

Styres af den boolske parameter deep_thinking. Når den er aktiveret, udfører modellen yderligere ræsonnering før generering.

Hvornår skal den aktiveres

Effekt på ydeevne

Deep thinking tilføjer ca. 3-5 sekunder, men koster nul ekstra credits.

Fejlhåndtering

Fejlsvarformat

{
  "error": {
    "code": "insufficient_credits",
    "message": "Not enough credits to complete this generation. Required: 7, Available: 3",
    "status": 402
  }
}

Fejlkoder

Gentagelsesstrategi

import time

def generate_with_retry(prompt, max_retries=3):
    for attempt in range(max_retries):
        try:
            response = requests.post(
                f"{BASE_URL}/images/generate",
                headers=headers,
                json={
                    "model": "seedream-5.0-lite",
                    "prompt": prompt
                }
            )

            if response.status_code == 429:
                retry_after = int(response.headers.get('Retry-After', 5))
                time.sleep(retry_after)
                continue

            response.raise_for_status()
            return response.json()

        except requests.exceptions.RequestException as e:
            if attempt == max_retries - 1:
                raise
            time.sleep(2 ** attempt)  # Eksponentiel backoff

Hastighedsbegrænsninger og bedste praksis

Hastighedsbegrænsninger pr. niveau

Bedste praksis

1. Brug webhooks frem for polling - mere effektivt
2. Batch når det er muligt - én batch-anmodning slår 50 individuelle anmodninger
3. Eksponentiel backoff - håndtér hastighedsbegrænsninger elegant
4. Cache resultater - gem billed-URL'er og metadata i din database
5. Download hurtigt - billed-URL'er udløber efter 24 timer
6. Overvåg kreditsaldo - tjek credits_remaining for at undgå afbrydelser
7. Brug metadata - tag genereringer med projekt-ID'er til sporing
8. Håndtér fejl elegant - ikke enhver generering lykkes

SDK-installation

Python:

pip install seedance

from seedance import SeedanceClient

client = SeedanceClient(api_key="your_key")
result = client.images.generate(
    model="seedream-5.0-lite",
    prompt="A beautiful sunset",
    deep_thinking=True
)

JavaScript:

npm install @seedance/sdk

import { SeedanceClient } from '@seedance/sdk';

const client = new SeedanceClient({ apiKey: 'your_key' });
const result = await client.images.generate({
    model: 'seedream-5.0-lite',
    prompt: 'A beautiful sunset',
    deepThinking: true
});

Almindelige integrationsmønstre

CMS-integration

Generér fremhævede billeder automatisk ved oprettelse af indlæg:

@app.route('/cms/webhook/new-post', methods=['POST'])
def handle_new_post():
    post = request.json

    result = generate_image(
        prompt=f"Blog header image for article about {post['title']}, "
               f"professional editorial style, 16:9",
        aspect_ratio="16:9",
        deep_thinking=True,
        metadata={"post_id": post["id"]}
    )

    update_post_featured_image(post["id"], result["image_url"])

E-handels-produktbilleder

def generate_product_images(product):
    prompts = [
        f"Product photo of {product.name}, white background, studio lighting, 1:1",
        f"Lifestyle photo of {product.name} in use, natural setting, 16:9",
        f"Product detail close-up of {product.name}, macro photography, 1:1"
    ]

    batch = client.images.batch_generate(
        generations=[
            {"model": "seedream-5.0-lite", "prompt": p}
            for p in prompts
        ]
    )

    return batch

Automatisering af sociale medier

def generate_weekly_social_content(brand, topics):
    generations = []
    for topic in topics:
        generations.append({
            "model": "seedream-5.0-lite",
            "prompt": f"{brand.style_prefix} {topic}, social media post, 1:1",
            "aspect_ratio": "1:1",
            "deep_thinking": True,
            "metadata": {"topic": topic, "platform": "instagram"}
        })

    return client.images.batch_generate(generations=generations)

---

Seedream 5.0 Lite API er den hurtigste vej fra tekstprompt til genereret billede med fuld understøttelse af deep thinking, batch-generering og webhook-integration.

Hent din API-nøgle → - 10 gratis credits, generér din nøgle under Indstillinger > API-nøgler, og kom i gang på få minutter.