Primary use stated or inferred cautiously from official documentation.
Seedream 5.5
Image generation and editing model introduced on November 9. Combines successive editing with RGBA transparency, layer decomposition and OCR for regular text. Available through BytePlus ModelArk, subject to account authorization.
The essential
Maximum input capacity when published by the source.
Documented maximum generation limit.
Verified availability channels.
Primary functional family and verified specialties.
Declared terms for API access, model weights, or self-hosting.
Limits and integration
| API ID | dola-seedream-5-5-261028 |
|---|---|
| Model type | Image |
| Access model | Paid API · early access depending on account |
| License | Proprietary |
| Deployment | Hosted API |
| Release | 09 NOV 2026 |
| Knowledge cutoff | Not published |
| Entrance | Text · Image |
| Exit | Image |
| Context window | Up to 10 reference images |
| maximum output | 1K, 1.5K or 2K · PNG or JPEG · up to 99 layers plus the base |
| Reasoning | Not published |
| Published tools | |
| Structured Outings | Not published |
| Batch processing | No sequential generation of image groups |
| Prompt cache | Not published |
| Fine-tuning | Not published |
| Verified platforms | BytePlus ModelArk |
Prepare editable compositions and transparent assets from text or up to ten images. Generates one image per request; decomposition can return a base and up to 99 layers. Optional OCR returns text, position and approximate styles for regular text layers.
Availability depends on account authorization and model activation. Use PNG for transparency: JPEG is opaque. Sequential group generation and streaming are not supported. OCR does not cover artistic text, handwriting, non-text layers or the base image. Check rights, readability and consistency after every edit.
What it can do
Orientation
Prepare editable compositions and transparent assets from text or up to ten images. Generates one image per request; decomposition can return a base and up to 99 layers. Optional OCR returns text, position and approximate styles for regular text layers.
Context
Up to 10 reference images · 1K, 1.5K or 2K · PNG or JPEG · up to 99 layers plus the base
Tools and integration
Not published
Access
BytePlus ModelArk · ap-southeast-1
Documented cost
| Concept | Worth | Unit/condition |
|---|---|---|
| Standard input | USD 0.008 / input image | Standard API rate |
| Cached input | Not published | Reading reused prefixes |
| Cache write or storage | Not published | The condition varies by provider |
| Standard output | USD 0.020–0.035 / image or layer, depending on pixels | May include reasoning tokens |
| Batch input | Not published | Asynchronous processing |
| Batch output | Not published | Asynchronous processing |
BytePlus charges USD 0.008 for each input image. Each output costs USD 0.020 up to 2,601,123.84 pixels or USD 0.035 above that, up to 4,624,220.16 pixels. In decomposition, each generated layer is billed by its dimensions. These are image prices, not per-million-token rates; check the exact tier before budgeting.
i Prices change and may depend on level, region or context length. Check the source before making a decision.
How to read the results
Continuous-editing demonstrations and provider comparisons do not guarantee the same result for every image or language. No independent numerical score has been verified for the cataloged API identifier. Check composition, preservation of unedited regions and text on representative material.
i We have not found published benchmarks for this model.
Inferama only highlights a “best result” when the metric, test set, configuration, and date allow for an equivalent comparison. The supplier's figures are presented as claims from its own source.
Chronology
dola-seedream-5-5-261028
Original November 9 announcement; API identifier and prices verified on November 15. The identifier suffix is not used as a release date.