What is Stable Diffusion?

Stable Diffusion is a family of image-generation models. Many releases offer downloadable weights for local workflows, while hosted services and third-party applications provide other access routes.

Which SD 3.5 model should I use?

Large prioritizes quality, Large Turbo uses four-step distilled generation, and Medium is a lighter local option. Compare them using your actual brief and compatible workflow.

What hardware does Stable Diffusion need?

Requirements depend on the checkpoint, precision, quantization, image size, batch size, and offloading. There is no universal VRAM minimum for every workflow.

Are LoRA and ControlNet the same?

No. LoRA adapters modify model behavior, often for a style or subject. ControlNet supplies structural guidance such as edges or depth. Both must be compatible with the selected model.

Does a fixed seed guarantee identical images?

No. A fixed seed helps controlled comparisons, but hardware, software versions, random-number generators, and pipeline settings can change results. Save the complete workflow and environment.

Can SD 3.5 be used commercially?

Covered models allow commercial use under the Community License for individuals or organizations below $1 million in total annual revenue, subject to its terms. Review enterprise requirements when reaching that threshold and check all derivative licenses.

Is local Stable Diffusion free to run?

Eligible model use may have no license fee, but local inference still requires hardware, electricity, storage, and maintenance. Hosted services have their own costs and terms.

Does Stability AI offer a hosted image API?

Yes. Its developer platform provides Stable Image services. Check the specific endpoint and terms rather than assuming it matches a local SD 3.5 checkpoint.

AI / Image Generation / Stability AI

Stable Diffusion

Open and API-accessible image generation family centered on Stable Diffusion 3.5, with extensive customization and Stability AI editing services.

Image Generation & EditingCompare this model Visit website

Starting Price

Self-host: Free under Community License for eligible users · API from $0.025/image (SD 3.5 Flash)

Maximum Resolution

SD 3.5 Large: ~1 MP · SD 3.5 Medium: 0.25–2 MP · Stable Image Ultra API: 1 MP native · dedicated upscalers output up to 4K

Image Editing

Stability AI API: Erase · Inpaint · Outpaint · Search & Replace · Recolor · Replace Background & Relight

Best For

Developers / Creators / Fine-tuning / Local deployment / Marketing / Production

About Stable Diffusion

Stable Diffusion 3.5 includes Large, Large Turbo, Medium and Flash variants. The ecosystem supports self-hosting, fine-tuning and LoRA workflows, while Stability AI API adds image-to-image, editing, control and 4K upscaling services.

Specifications

Pricing

3 specifications

Starting Price	Self-host: Free under Community License for eligible users · API from $0.025/image (SD 3.5 Flash)
Free Plan	Community License: free for non-commercial use and commercial use below $1M annual revenue · Developer API starts with 25 free credits
Usage Limits / Credits	API: 1 credit = $0.01 · SD 3.5 Flash 2.5 credits · Medium 3.5 · Large Turbo 4 · Large 6.5 · Ultra 8 per image

Overview

3 specifications

Primary Use Case	Customizable AI image generation · self-hosting · production image workflows
Best For	Developers / Creators / Fine-tuning / Local deployment / Marketing / Production
Platforms	Self-hosted/local · Stability AI API · Stable Assistant · cloud partners

Image Generation

6 specifications

Text-to-Image	Stable Diffusion 3.5 Large / Large Turbo / Medium / Flash text-to-image
Image-to-Image	Stable Image generation endpoints accept a starting image with adjustable strength; ecosystem also supports custom img2img workflows
Maximum Resolution	SD 3.5 Large: ~1 MP · SD 3.5 Medium: 0.25–2 MP · Stable Image Ultra API: 1 MP native · dedicated upscalers output up to 4K

Image Editing	Stability AI API: Erase · Inpaint · Outpaint · Search & Replace · Recolor · Replace Background & Relight
Inpainting	Dedicated Stability AI Inpaint API + customizable self-hosted workflows
Outpainting	Dedicated Stability AI Outpaint API
Object Removal	Dedicated Erase Object API
Background Removal	Dedicated Remove Background API
Background Replacement	Replace Background & Relight API
Generative Fill	Inpaint / Outpaint services for masked and extended generative content
Upscaling	Creative and Conservative Upscalers to 4K · Fast Upscaler 4× up to 4 MP

Reference Images	Image-to-image + Control Style/Structure/Sketch services; self-hosted reference workflows depend on implementation
Style Control	Control Style · Style Transfer · self-hosted LoRA/fine-tuning and prompt-based style control
Character Consistency	No single native SD 3.5 character-lock feature · commonly implemented through fine-tunes, LoRAs and reference/control workflows
Style Reference	Control Style / Style Transfer API plus self-hosted reference and adapter workflows
Pose / Composition Control	Control Sketch / Structure APIs · advanced pose controls available through ecosystem implementations
Seed Control	Seed parameter supported in Stability AI generation API and standard self-hosted diffusion workflows

Photorealism	Excellent with Stable Image Ultra / SD 3.5 Large; quality varies by model and fine-tune
Text Rendering	Improved typography and prompt adherence in SD 3.5; Ultra is optimized for stronger typography
Typography / Logo Generation	Raster typography/design generation supported; quality depends on model and prompt
Vector Generation	No native SVG/vector output in core Stable Diffusion 3.5 image models
Transparent Background	Remove Background API supports isolated foreground workflows; native transparent generation is not a core SD 3.5 base-model feature

Available Models	Stable Diffusion 3.5 Large · Large Turbo · Medium · Flash · Stable Image Ultra/Core · legacy SDXL
Model Selection	Choose SD 3.5 variant by quality/speed/hardware needs; Stable Image services provide optimized hosted alternatives
Custom Models	Highly customizable · supports community fine-tunes, LoRAs and derivative workflows under applicable license terms
Model Training	Fine-tuning and LoRA workflows supported; Stability AI Enterprise also offers custom model training upgrades

API Access	Stability AI REST v2beta · Generate / Edit / Control / Upscale services
Integrations	Self-hosted ecosystem · Stability AI API · Stable Assistant · cloud partners
Automation Support	REST API with asynchronous polling support · self-hosted pipelines can be automated

Stable Diffusion

About Stable Diffusion

Specifications

Pricing

Overview

Image Generation

Image Editing

Control & Consistency

Quality & Design

Models & Customization

Developer & Integration

Commercial & Business

Compare Stable Diffusion

Compare before you choose

Similar AI models

FLUX — Black Forest Labs

Leonardo.Ai

Recraft

Google Nano Banana

Adobe Firefly

Ideogram

Krea AI

Photoroom

Reviews & Guides

Stable Diffusion: SD 3.5, Local Workflows & Guide (2026)

Commercial Use	Community License free for organizations under $1M annual revenue; Enterprise License required above $1M for covered Core Models
Team Plan	No standard seat-based Stable Diffusion team plan · organizations can use API credits or self-host under applicable license
Enterprise Plan	Enterprise License with commercial deployment, implementation support and optional custom model training