bytedance/seedvr/upscale/image

Name: SeedVR2
Author: Bytedance

Seedvr Upscale Image by Bytedance - AI-powered image editing, style transfer, and transformation. Edit photos with natural language instructions, remove backgrounds, change styles, and enhance images effortlessly.

Input

Image URL or upload(string)

Upload

PNG, JPEG, WebP, or GIF · 20 MiB maximum

The input image to be processed

This field is required.

Target Resolution(string)

The target resolution to upscale to. Allowed values: 2k, 4k.

Output Format(string)

The format of the output image. Allowed values: png, jpg.

Idle

Example output — click Run to generate your own

Production endpoint

Send your first request

OpenAI-compatible endpoint with unified authentication and usage tracking.

POSThttps://api.sandbase.ai/v1/run

Model IDbytedance/seedvr/upscale/image

Input Schema

3 parameters · 1 required · 2 optional

Parameter	Type	Required	Description
`image`	string	Required	The input image to be processed
`output_format`	string	Optional	The format of the output image. · Options: png, jpg · Default: "jpg" pngjpg
`target_resolution`	string	Optional	The target resolution to upscale to. · Options: 2k, 4k · Default: "4k" 2k4k

Output Schema

Field	Type	Description
`id`	string	Unique identifier for the generation task
`status`	string	Task status: pending, running, completed, failed, timeout
`model`	string	Model used for the generation
`outputs`	array	Array of output items
`outputs[].url`	string	URL of the generated artifact
`outputs[].content_type`	string	MIME type (e.g. image/png, video/mp4)
`error`	object \| null	Error details if failed, null on success
`error.type`	string	Machine-readable error type code
`error.message`	string	Human-readable error description

Async Workflow

This model uses asynchronous execution. Submit a request and poll for the result.

Submit — POST to /v1/run, receive an id
Poll — GET /v1/run/{id} until status is completed, failed, or timeout
Retrieve — Read outputs from the completed response

Code Examples

Ready-to-run snippets

# Step 1: Submit
curl -X POST https://api.sandbase.ai/v1/run \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
  "model": "bytedance/seedvr/upscale/image",
  "image": "https://static.sandbase.ai/examples/bytedance/seedvr/upscale/image/input_image_0.png",
  "output_format": "jpg",
  "target_resolution": "4k",
  "prompt": "a beautiful sunset over mountains"
}'

# Step 2: Poll result (replace <id>)
curl https://api.sandbase.ai/v1/run/<id> \
  -H "Authorization: Bearer YOUR_API_KEY"

← Back to SeedVR2

API README

SeedVR2 Image Upscaler

SeedVR2 Image Upscaler restores and enlarges still images through a dedicated image workflow. It is built for cases where the visual content is already established but the delivered asset needs more usable resolution. The result is returned as a processed image suitable for downstream publishing or archiving.

The route exposes a focused choice between 2K and 4K output, with JPEG or PNG delivery. This keeps the request compact while making the target size and file type explicit. Teams can place it after image generation, photography, or asset ingestion without redesigning the original composition.

Highlights

Dedicated SeedVR2 image upscaling. The route applies SeedVR2 specifically to still-image enlargement. It gives image assets a direct restoration path instead of routing them through a video workflow.
Single-step restoration architecture. SeedVR2 is presented as a one-step large diffusion-transformer restoration model. Its restoration process does not depend on an additional diffusion prior.
High-resolution attention design. The SeedVR2 project describes adaptive window attention for high-resolution restoration. That design targets demanding image sizes while keeping the restoration operation unified.
Strong one-step restoration performance. The published project reports comparable or better restoration performance in a single step. This makes the model notable for combining a streamlined process with high-quality recovery.

Pricing

Billing unit	Price
Base price per request	$0.001

When to Use

✅ Good fit	❌ Consider alternatives
Enlarging a generated image before high-resolution delivery	Creating a completely new composition from text
Restoring a low-resolution still for a sharper digital asset	Processing motion footage that needs frame continuity
Preparing a 2K or 4K image for publishing	Requiring an exact custom pixel size outside the available targets
Converting the restored result to JPEG or PNG	Needing an editable vector or layered design file
Standardizing mixed still assets to a common delivery tier	Preserving every original pixel without restoration changes

Prompt Guide

This is an input-guided restoration route rather than a text-prompted generator. Supply one accessible image, then choose the target resolution and delivery format.

Image: <publicly accessible image URL>
Target resolution: <2k or 4k>
Output format: <jpg or png>

Example request:

{
  "image": "https://example.com/input-image.jpg",
  "target_resolution": "4k",
  "output_format": "jpg"
}

Technical Specs

Spec	Value
Model ID	`bytedance/seedvr/upscale/image`
Required input	`image`
Target resolution	`2k` or `4k`; default `4k`
Output format	`jpg` or `png`; default `jpg`
Output	Image URL and content type
Execution	Asynchronous job

SeedVR2 Video Upscaler — use the sibling route when the asset is video and temporal consistency matters.