> ## Documentation Index
> Fetch the complete documentation index at: https://docs.snapgen.org/llms.txt
> Use this file to discover all available pages before exploring further.

# Seedream 5.5

> Generate or edit one image with Seedream 5.5 and up to ten references.

Generate an image from a text prompt, or add public image URLs to guide an edit.
The route returns one image at `1K`, `1.5K` or `2K`.

| Property | Value |
| - | - |
| Public model ID | `seedream-5.5` |
| Output count | One image per request (`n: 1`) |
| Base output price | \$0.02 at 1K or 1.5K; \$0.035 at 2K |
| Reference images | Up to ten public HTTP(S) URLs; \$0.008 per reference |
| Output formats | PNG or JPEG |

The total is the selected base price plus the price of every supplied reference.
For example, a 1K output with two references costs \$0.036. Check the
[live model page](https://snapgen.org/models/seedream-5-5) for current prices.

**model** `string` *required*

Use `seedream-5.5`.

**prompt** `string` *required*

A non-empty creative brief of at most 7,000 characters.

**size** `string` (default: auto)

`auto`, `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`, `2:1`
or `1:2`. Custom pixel sizes are not accepted.

**resolution** `string` (default: 1K)

`1K`, `1.5K` or `2K`. The selected tier determines the base output price.

**image\_urls** `string[]`

Optional: up to ten public HTTP(S) image URLs. Each reference is billed.

**background** `string` (default: auto)

`auto`, `opaque` or `transparent`. Transparent output requires PNG. Describe
the isolated subject and review the resulting alpha channel.

**output\_format** `string` (default: png)

`png` or `jpeg`. Transparent output requires `png`.

**n** `integer` (default: 1)

Only `1` is supported.

## Generate an image

```json theme={null}
{
  "model": "seedream-5.5",
  "prompt": "A ceramic cup on a sunlit table, editorial product photography",
  "size": "1:1",
  "resolution": "1K",
  "n": 1
}
```

## Edit with references

```json theme={null}
{
  "model": "seedream-5.5",
  "prompt": "Keep the product shape, and change the background to pale linen",
  "size": "1:1",
  "resolution": "1K",
  "n": 1,
  "image_urls": ["https://example.com/product.png"]
}
```

Send the JSON body to `POST /v1/images/generations`. Use `Prefer: respond-async`
to receive a task ID immediately, then poll `GET /v1/tasks/{id}` with the same
API key. A waiting request may also return a task if it has not finished.
Keep the same idempotency key when retrying an uncertain submission; download
temporary result URLs promptly.

This integration does not expose custom pixel dimensions, batch output or layer
decomposition.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.