Bytedance Seedream V4 Edit Sequential

Bytedance Seedream V4 Edit Sequential

Playground

Seedream 4.0 is a state-of-art image model by Bytedance. Seedream 4.0: Surpassing nano banana
in every aspect.

Features

Seedream 4.0 is a state-of-art image model by Bytedance. Seedream 4.0: Surpassing nano banana in every aspect.

Multi-modal image generation support: Seedream 4.0 is the first to support multi-modal image generation, enabling text-to-image, image-editing, and group image generation with a single model.

Outstanding model advantages: It features five major highlights: precise instruction editing, high feature retention, deep understanding ability, ultra-fast inference speed, and ultra-high-resolution output. For example, it takes as little as 1.8 seconds to generate a 2K image in text-to-image mode.

Diverse scene applications: Covers various scenarios such as commercial design (including posters, brand clothing, packaging, e-commerce, etc.), entertainment (such as anime and movie roles), fine art, architectural design, and more. It supports various complex editing operations, such as object addition and deletion, attribute change, style change, structural adjustment (e.g., face swapping), etc.

Prompt writing guide: Use clear and specific instructions following the formula of “change action + change object + target feature.” While generating multiple images, use words like “a series of,” “group of images” to maintain consistency.

Text-to-image writing guide: The image content in coherent natural language and group images for a specific scene description. Use professional vocabulary in its original language and special image terms to match the scene description.

Problem self-check list: Identify common problems such as incorrect instructions, inadequate description, missing details, or redundant instructions. It recommends rechecking image editing objects, and insufficient aesthetic effects, and enhancing background consistency.

Rich editing functions: It includes style/image editing (element addition and deletion, frame/brush editing, style transformation, texture replacement, etc.), structural editing (object structure modification, object reshaping, etc.), and feature editing (color tone, detail transformation, attribute adjustment, etc.).

# Submit the task
curl --location --request POST "https://api.wavespeed.ai/api/v3/bytedance/seedream-v4/edit-sequential" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}" \
--data-raw '{
    "size": "2048*2048",
    "max_images": 1,
    "enable_base64_output": false,
    "enable_sync_mode": false
}'

# Get the result
curl --location --request GET "https://api.wavespeed.ai/api/v3/predictions/${requestId}/result" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}"