Seedream - 5.5 Layer Decomposition

Seedream
Seedream on FuseAITools is ByteDance's image suite: 5 Lite at 2K/3K, 5 Pro (1K/2K), 5 Flash (1K/1.5K/2K flat), and the latest 5.5 tier — all with up to 10 reference images, background control, and layer decomposition that splits artwork into editable PNG layers. Resolution is the billing dimension. Pair with Seedance to turn stills into video clips.
Configuration
No image generated yet
Fill in the form and click "Generate Image" to start.
🧩 Seedream 5.5 Layer Decomposition
Send one source image and get back a base plus separated PNG layers. The prompt is optional (≤2000 chars, with <bbox>x1 y1 x2 y2</bbox> targeting), you can cap output with Max Images (2–100) and toggle text recognition.
What is Seedream 5.5 Layer Decomposition?
Seedream 5.5 Layer Decomposition on FuseAITools (seedream-5-5-layer-decomposition) separates one image into reusable layers. Required: one imageUrl (JPEG/PNG/WebP ≤10MB) and a resolution of 1K / 1.5K / 2K. Optional prompt (≤2000, with <bbox>), maxImages (2–100) and textRecognition. Layers are always PNG.
🌟 Seedream 5.5 on FuseAITools
Seedream 5.5 on FuseAITools is ByteDance's newest image tier: 5000-character prompts, 8 aspect ratios, a 10-image reference capacity and layer decomposition—now with explicit background control (auto / transparent / opaque). Unlike 5 Flash's flat rate, 5.5 is billed by resolution: 1K and 1.5K share a price while 2K costs more. Three workflows—text to image, image to image and layer decomposition—cover the full pipeline. New users receive 10 free credits on sign-up.
✨ Seedream 5.5 Core Features
The newest Seedream generation, sitting above 5 Pro and 5 Flash in capability.
1K / 1.5K share a price; 2K is higher—pay precisely for the detail you need.
Required prompt of 5 to 5000 characters on generation and editing.
Pick auto, transparent or opaque—transparent output requires PNG.
Composite or restyle 1–10 images (JPEG/PNG/WebP, each ≤10MB).
Layer decomposition adds Max Images (2–100) and text recognition, with optional <bbox> targeting.
🎯 Built for These Scenarios
⚙️ Seedream 5.5 Technical Specs
Parameters match the FuseAITools Seedream form and API (seedream-5-5-* models).
| Workflow | Input | Resolution (billing) | Key controls |
|---|---|---|---|
| 5.5 Text to Image | Prompt 5–5000 chars | 1K / 1.5K / 2K — 2K higher | 8 aspect ratios, background, PNG/JPEG, NSFW filter |
| 5.5 Image to Image | 1–10 image URLs + prompt 5–5000 | 1K / 1.5K / 2K — 2K higher | Optional aspect (incl. auto), background, PNG/JPEG; each ≤10MB |
| 5.5 Layer Decomposition | Exactly 1 source image + optional prompt ≤2000 | 1K / 1.5K / 2K — 2K higher | <bbox> targeting, Max Images 2–100, text recognition, layers always PNG |
📊 Seedream Generations Compared: 5 Lite · 5 Pro · 5 Flash · 5.5
| Dimension | 5 Lite | 5 Pro | 5 Flash | 5.5 |
|---|---|---|---|---|
| Positioning | Multimodal workhorse | Premium quality | Speed-first lane | Latest tier |
| Prompt limit | 2995 / 2996 | 4–5000 | 4–5000 | 5–5000 |
| Resolution / billing | Basic 2K / High 3K, by ratio + quality | 1K / 2K (layers 1K–2K) | 1K / 1.5K / 2K at one price | 1K / 1.5K / 2K, 2K higher |
| Reference images | 1–5, each ≤10MB | 1–10, each ≤30MB | 1–10, each ≤30MB | 1–10, each ≤10MB |
| Background control | — | — | — | auto / transparent / opaque |
| Layer decomposition | — | Yes | Yes | Yes + Max Images / text |
| Aspect ratios | 8 (1:1 → 21:9) | 8 (1:1 → 21:9) | 8 (1:1 → 21:9) | 8 + auto (i2i) |
| Best for | Routine 2K/3K stills | Hero art and detail | Bulk drafts at flat cost | Latest quality with resolution billing |
❓ FAQ (Seedream 5.5)
Seedream 5.5 — Three Workflows
Choose the mode that matches your task:
5.5 Text to Image
Seedream 5.5 Text to Image
Latest prompt-to-image—5000 chars, 8 ratios, background, 1K/1.5K/2K.
Open this workflow →5.5 Image to Image
Seedream 5.5 Image to Image
Remix up to 10 references; optional aspect with auto support.
Open this workflow →5.5 Layer Decomposition
Seedream 5.5 Layer Decomposition
Layer splits with Max Images and text recognition.
Current workflow