What is SDXL Image-to-Image?

SDXL Image-to-Image uses the same 3.5B parameter SDXL architecture but takes an existing image as input along with a text prompt. The strength parameter (0.05–1.0) controls how much the output departs from the original — low values make subtle adjustments, high values reimagine the image entirely.

How much does SDXL Image-to-Image cost on Runflow?

Same pricing as standard SDXL — approximately $0.003 per image at $0.00125/compute second. Cold start time is not billed.

What is the strength parameter?

Strength controls the balance between the input image and the generated output. At 0.05, the output closely resembles the input with minimal changes. At 1.0, the model essentially ignores the input and generates from scratch using only the prompt. The default of 0.95 allows significant creative freedom while maintaining some structural guidance from the input.

Can I use LoRA weights with image-to-image?

Yes. The full SDXL LoRA ecosystem (50,000+ community models) works with image-to-image generation. Apply style LoRAs for consistent visual transfer across images.

Can I use SDXL Image-to-Image commercially?

Yes. Same CreativeML Open RAIL++-M license as standard SDXL — commercial use with no revenue threshold.

Do I need to manage GPUs?

No. Runflow handles all infrastructure, including GPU provisioning, scaling, and failover. You call our API and receive results — no DevOps required.

Stability AI

SDXL Image-to-Image API

SDXL's image-to-image variant — transform existing images with text prompts and adjustable strength control. Full LoRA and embedding support with the same 50,000+ community ecosystem.

sdxl-image-to-image3.5B paramsUNet-based latent diffusion with dual CLIP text encoders

Updated March 2026·Early access

Run the API How Sentinel controls quality

Price

~$0.003/image

Usage-based, per compute second

Latency

~2–4s

25 steps at 1024×1024

Cold Start

5–10s

Not billed · handled by Runflow

SDXL family by Stability AI

Explore all variants in the SDXL family available on Runflow.

SDXL

Standard

Standard 3.5B text-to-image generation

SDXL Image-to-Image

Image-to-Image

Image-guided generation with strength control

SDXL Inpainting

Inpainting

Mask-based region editing

SDXL Lightning

Fast

1–4 step generation, sub-second latency

SDXL ControlNet Union

ControlNet

10+ ControlNet types in one model

SDXL Fooocus

Enhanced

Enhanced quality with automated prompt optimization

Benchmarked for your use case

Every model on Runflow is scored per vertical — not generic leaderboard numbers. Scores come from domain-expert evaluation on real production prompts.

Use Case	Assessment	Notes
Portrait	Pending	Portrait editing quality
Fashion	Pending	Fashion editing quality
Product	Pending	Product editing quality
Creative	Pending	Creative editing quality

API Parameters

Full parameter reference for the sdxl-image-to-image model endpoint.

Parameter	Type	Required	Description
prompt	string	required	Text description of the desired output. Works with both natural language and keyword-style prompts.
image_url	string	required	URL of the input image. Accepts JPEG, PNG, WebP, GIF, AVIF.
strength	float	optional	How much the generated image departs from the input. 0.05 = subtle; 1.0 = full regeneration. Default: 0.95.
negative_prompt	string	optional	Highly effective on SDXL — describe what to exclude from the output.
image_size	object \| string	optional	Output dimensions. Presets: square_hd, square, landscape_4_3, landscape_16_9, portrait_4_3, portrait_16_9. Or custom {width, height}.
num_images	integer	optional	Number of images to generate. Range: 1–8. Default: 1.
num_inference_steps	integer	optional	Denoising steps. 20–25 for standard, 30–50 for max fidelity. Default: 25.
guidance_scale	number	optional	Classifier Free Guidance. 7–9 recommended. Default: 7.5.
seed	integer	optional	Reproducibility seed. Same seed + same parameters = identical output.
loras	array[object]	optional	Array of LoRA weights to apply. Stack multiple LoRAs for combined effects.
embeddings	array[object]	optional	Textual inversion embeddings to apply during generation.
preserve_aspect_ratio	boolean	optional	Maintain input image aspect ratio. Default: false.
expand_prompt	boolean	optional	Auto-expand prompt with additional detail via LLM. Default: false.
format	string	optional	Output image format. Options: "jpeg", "png". Default: "jpeg".
enable_safety_checker	boolean	optional	Enable content safety filter. Default: true.
sync_mode	boolean	optional	If true, returns the image as a data URI.

Technical Specifications

ArchitectureUNet-based latent diffusion (3.5B base UNet, dual CLIP text encoders)

Parameters3.5B (base); ~6.6B with refiner

Input FormatsJPEG, PNG, WebP, GIF, AVIF

Max Resolution1024 × 1024 native

Strength Range0.05–1.0 (controls departure from input)

VRAM7–8 GB (FP16)

Output FormatJPEG, PNG

LoRA SupportYes — 50,000+ community LoRAs on Civitai

LicenseCreativeML Open RAIL++-M (no revenue restrictions)

Why run SDXL Image-to-Image on the Runflow API

Flux AI models are available on many platforms. Here's what makes the Runflow image generation API different.

Adjustable strength control

Fine-tune how much the output deviates from the input image — from subtle tweaks (0.05) to complete reimagining (1.0) — all with a single parameter.

Full LoRA ecosystem

Apply any of 50,000+ community LoRAs during image-to-image generation for style transfer, character consistency, and domain adaptation.

Effective negative prompts

Unlike newer models, SDXL's negative prompt system is highly responsive — giving precise control over what to exclude from edited outputs.

Aspect ratio preservation

Optional preserve_aspect_ratio flag maintains the input image's proportions — no manual calculation or cropping needed.

Battle-tested pipeline

The same proven SDXL architecture used in millions of production workflows, now applied to image-guided generation.

Explore other models

Same API, different strengths. Switch models with one parameter change.

Black Forest Labs

Frequently asked questions

Resources

Sentinel Quality Scoring

Learn how Sentinel evaluates every output for quality before delivery.

Start generating images with SDXL Image-to-Image

Get API access in minutes. No GPU setup, no infrastructure to manage. Pair with Sentinel to control the quality of every AI-generated image you ship.

Run the API Talk to Founders