Exact Doubao Model ID
Use doubao-seedream-5-5-261028, the Volcengine Ark identifier for Seedream 5.5. It names a deployment route, not a separate image model.
Quick Start is not available for this model yet.
Start from the cost of a single sample and pick a testing budget. GPTProto rates are 15% below list price.
| Scenario | Bytedance list | OpenRouter | GPTProto | You save / mo |
|---|---|---|---|---|
| Quick tests & prototypes100 images / mo | $2.80 | $2.95 | $2.38 | −$0.42≈ $5.04 / yr |
| Product & e-commerce photos1,000 images / mo | $28.00 | $29.54 | $23.80 | −$4.20≈ $50.40 / yr |
| Ad / marketing creatives2,500 images / mo | $70.00 | $73.85 | $59.50 | −$10.50≈ $126.00 / yr |
| Social media content10,000 images / mo | $280.00 | $295.40 | $238.00 | −$42.00≈ $504.00 / yr |
Use the Doubao Seedream 5.5 API for multi-turn image editing, native transparent output, multi-reference composition, and fine-grained layer decomposition. The exact upstream model ID is doubao-seedream-5-5-261028. This is not a separate “Doubao edition” with different image quality, speed, or features. It is the Volcengine Ark name for the same underlying Seedream 5.5 model that BytePlus lists as dola-seedream-5-5-261028. GPTProto keeps the functions and price the same, so the practical difference is the identifier and route shown in the API Usage block.
Exact Doubao Model ID
Use doubao-seedream-5-5-261028, the Volcengine Ark identifier for Seedream 5.5. It names a deployment route, not a separate image model.
Doubao Seedream 5.5 is ByteDance's Seedream 5.5 image generation and editing model as published in the Volcengine Ark catalog. Volcengine's official documentation calls it doubao-seedream-5-5-261028, while BytePlus ModelArk uses dola-seedream-5-5-261028 for its international catalog.
The prefix identifies the upstream service catalog; it does not describe a new architecture or a different checkpoint. Both IDs refer to Seedream 5.5 and expose the same core workflows: text-to-image, single-image editing, multi-reference editing with up to 10 images, native RGBA generation, and decomposition into one base image plus as many as 99 layers.
On GPTProto, the Doubao and Dola listings use the same account balance and have the same price. Use this page when your application, catalog, or migration checklist refers to the doubao- model string. For the full capability guide, prompt examples, and comparisons with other image models, see the Seedream 5.5 API page.
| Field | Doubao Seedream 5.5 | Seedream 5.5 / Dola listing |
|---|---|---|
| Official upstream catalog | Volcengine Ark | BytePlus ModelArk |
| Official model ID | doubao-seedream-5-5-261028 |
dola-seedream-5-5-261028 |
| Underlying model | Seedream 5.5 | Seedream 5.5 |
| Image capabilities | Same | Same |
| Output tiers | 1K, 1.5K, and 2K | 1K, 1.5K, and 2K |
| Reference images | Up to 10 | Up to 10 |
| Native RGBA | Supported | Supported |
| Layer decomposition | One base image plus up to 99 layers | One base image plus up to 99 layers |
| GPTProto pricing | Same live rate | Same live rate |
| GPTProto API key | Same account key | Same account key |
| Which one to choose | When the page or integration specifies the doubao- ID |
When the page or integration specifies the dola- ID |
The short answer is that Doubao Seedream 5.5 and Dola Seedream 5.5 are two provider-facing names for the same model. Do not compare them as if one were a faster, cheaper, or higher-quality variant. Compare the model strings and API routes instead.
| Workflow | What the Doubao model can do |
|---|---|
| Text-to-image | Generate one image from a prompt at 1K, 1.5K, or 2K. |
| Multi-reference editing | Combine or edit up to 10 reference images in one request. |
| Multi-turn editing | Pass the latest result into the next request and specify what to change and preserve. |
| Transparent output | Generate native RGBA images with full or partial transparency from text or image input. |
| Layer decomposition | Split one input image into a base image and up to 99 positioned PNG layers. |
| Rich-text OCR | Return recognized text content and approximate styling data for regular text layers. |
This summary is intentionally compact. The generic page should remain the main destination for searches about transparent PNG output, 99-layer decomposition, prompt writing, pricing comparisons, and Seedream 5.5 versus other image models.
Use the identifier displayed on the model page that your request is calling. If the GPTProto API Usage block shows doubao-seedream-5-5-261028, keep that exact spelling in the model field or route. Do not shorten it to doubao-seedream-5-5 unless the live documentation explicitly provides that alias.
When moving an existing workflow from the Dola listing, keep the same prompt, reference images, output size, transparency choice, and layer settings. Change only the confirmed model identifier or endpoint path shown by GPTProto. A Dola identifier copied into a Doubao-specific route—or the reverse—may return a model-not-found or routing error even though the underlying image model is the same.
Outside GPTProto, Volcengine Ark and BytePlus ModelArk use separate service endpoints, accounts, and credentials. A provider-issued key from one platform should not be assumed to work on the other. GPTProto removes that account split by using one GPTProto key for both listings and the rest of its supported model catalog.
Guides, comparisons, and updates related to this model.
All Articles
ByteDance's Seedance 2.5 Flash runs 2x faster at ~¥0.99/s (~$0.14/s) for 30s 720p clips—no 4K. See real trade-offs, Kling 4.0 Flash vs 2.5 specs & API code.

Seedream 5.5 brings more stable editing across 100 rounds, native RGBA output, and clearer small text. See what changed and what still needs testing.

See a cat sketch turn into a looping GIF with ChatGPT Images 2.6. Compare Flare and Sunburst and use either on GPTProto at 5% below OpenAI API rates.

We tested 6 AI image API routes across 18 UI, stock photo, and product outputs. Compare image quality, text accuracy, pricing, and developer fit.
Input
Output
