Category11free & open source
🖼️

Image Generation

Diffusion and flow-matching AI models for text-to-image, inpainting, outpainting, ControlNet composition, and high-resolution upscaling — including FLUX.1, Stable Diffusion 3, and DALL-E 3.

6AI Models14Tools
Most Popular In
OverviewPopularOpen Source
Notable Developers
Black Forest Labs (FLUX)Stability AI (SD3)OpenAI (DALL-E 3)IdeogramAdobe Firefly
Updated Jun 12, 2026
Curated by FreeAPIHub editors
Topics:Text-to-ImageInpainting & OutpaintingControlNet ConditioningStyle TransferHigh-Res UpscalingCommercial-Safe Generation
20 of 20
⭐Top Resources

ControlNet

Model · Stanford University (Lvmin Zhang)
Apache 2.0

ControlNet adds precise spatial control to Stable Diffusion: condition image generation on edges, depth, human pose, segmentation and more, so outputs follow a reference structure. Open-source under Apache 2.0.

↓ 1M+Not rated yetView

StyleGAN2

Model · NVIDIA Research
NVIDIA Source Code License

StyleGAN2 is NVIDIA's landmark generative adversarial network for ultra-realistic image synthesis, especially faces. Its style-based generator gives fine control over image attributes and set the standard for high-fidelity GANs.

↓ 1M+Not rated yetView

DreamBooth

Model · Google Research
Apache 2.0

DreamBooth is a fine-tuning technique for text-to-image models. Using just a few photos of a subject, it teaches a diffusion model to generate that specific person, pet or object in new scenes, styles and contexts.

↓ 1M+Not rated yetView

Pix2Pix

Model · UC Berkeley (Phillip Isola et al.)
BSD-2-Clause

Pix2Pix is a classic conditional GAN for image-to-image translation. Trained on paired images, it learns to turn one type of image into another — edges to photos, maps to satellite, sketches to renders — with a general framework.

↓ 920K+Not rated yetView

DALL·E Mini

Model · Boris Dayma & Community
Apache 2.0

DALL·E Mini (now Craiyon) is an open text-to-image model that brought AI image generation to the masses. Lightweight and free to use, it produces quirky, recognizable images from any prompt and sparked a viral creative craze.

↓ 3M+Not rated yetView

Stable Diffusion

Model · Stability AI
Open Source

Stable Diffusion is the landmark open text-to-image model from Stability AI. Using efficient latent diffusion, it generates high-quality images from text prompts on consumer GPUs and spawned a vast creative ecosystem.

↓ 42K+Not rated yetView
Canva AI logo

Canva AI

🔥 Hot
by Canva

Canva is a hugely popular design platform whose Magic Studio adds AI across the board: generate images and text, remove and replace backgrounds, resize designs, write copy and turn prompts into designs.

FreemiumView tool
Microsoft Designer logo

Microsoft Designer

🔥 Hot
by Microsoft

Microsoft Designer is a free AI design app for social posts, graphics and invitations. Generate images and on-brand designs from a prompt, with templates, and edit them in a simple Canva-style editor.

Leonardo AI logo

Leonardo AI

🔥 Hot
by Leonardo.Ai (Canva)

Leonardo AI is a creative image-generation platform for art, game assets and design. Generate and edit images with fine control, fine-tune your own models, and build on its API - popular with creators and studios.

FreemiumView tool
Ideogram logo

Ideogram

🔥 Hot
by Ideogram

Ideogram is an AI image generator known for accurate, legible text inside images. Generate posters, logos and graphics from prompts, with strong typography, style controls and an API for developers.

FreemiumView tool
Flux AI logo

Flux AI

🔥 Hot
by Black Forest Labs

FLUX is a family of top-tier image generation models from Black Forest Labs. Known for strong prompt-following and quality, it offers open-weight models to self-host and fast hosted APIs.

Open SourceView tool
Looka AI logo

Looka AI

🔥 Hot
by Looka

Looka is an AI logo maker and brand kit builder. Answer a few questions about your business and it generates logo options you can customise, then export logos and a full brand kit of matching assets.

FreemiumView tool
Upscayl logo

Upscayl

🔥 Hot
by Upscayl (open source)

Upscayl is a free, open-source desktop app that upscales and enhances images using AI. Turn low-resolution pictures into sharp, higher-resolution versions locally - cross-platform, with no subscription.

Open SourceView tool
Remove.bg logo

Remove.bg

🔥 Hot
by Kaleido (Canva)

remove.bg removes the background from any image automatically in seconds. Upload a photo and get a clean cutout with a transparent background - free for web previews, with an API and bulk tools for scale.

FreemiumView tool
Krita AI Diffusion logo

Krita AI Diffusion

🔥 Hot
by Acly (open source)

Krita AI Diffusion is a free, open-source plugin that brings Stable Diffusion into Krita. Generate, inpaint and refine images on a real painting canvas with layers, masks and full local control.

Open SourceView tool
Bing Image Creator logo

Bing Image Creator

🔥 Hot
by Microsoft

Bing Image Creator (now Microsoft Copilot's image generator) makes images from text prompts for free, powered by DALL-E. Generate art and concepts in the browser or inside Copilot, no subscription needed.

Playground AI logo

Playground AI

🔥 Hot
by Playground

Playground is an AI image tool for creating and editing graphics and art. Generate images from text, edit with AI, and design social posts, logos and product shots in a browser-based mixed canvas and editor.

FreemiumView tool
Vizcom AI logo

Vizcom AI

🔥 Hot
by Vizcom

Vizcom turns rough sketches into rendered product concepts in seconds. Aimed at industrial and product designers, it keeps your linework while adding material, lighting and colour with AI.

FreemiumView tool
Stable Diffusion logo

Stable Diffusion

🔥 Hot
by Stability AI

Stable Diffusion is the landmark open-weight image generation model from Stability AI. Generate and edit images from text, run it free on your own hardware, fine-tune it, or use it through hosted APIs.

Open SourceView tool
Adobe Firefly logo

Adobe Firefly

🔥 Hot
by Adobe

Adobe Firefly is Adobe's family of generative AI models for images, video and design. Trained on licensed and public-domain content for commercial safety, it powers standalone tools and Photoshop's AI features.

FreemiumView tool
Showing 20 of 20 resources

More to explore

Explore related categories

All categories
📈

Productivity

33 resources
2APIs31Tools
Browse category
🗣️

Natural Language Processing

30 resources
4APIs30Models
Browse category
💻

Development

26 resources
26APIs
Browse category
🧪

Science & Nature

14 resources
14APIs
Browse category

Learn more

From our blog

DeepSeek API Tutorial: Free, Low-Cost AI in Python (2026)
Tutorials

DeepSeek API Tutorial: Free, Low-Cost AI in Python (2026)

Jun 11, 2026 6 min read
Free Vector Database & Embeddings APIs in 2026
AI Tools

Free Vector Database & Embeddings APIs in 2026

Jun 11, 2026 5 min read
How to Build a Free MCP Server (Model Context Protocol)
Tutorials

How to Build a Free MCP Server (Model Context Protocol)

Jun 11, 2026 6 min read

About this category

Image Generation — developer guide

What Are Image Generation Models?

Image generation models create photorealistic photographs, artistic illustrations, architectural visualisations, product mockups, and virtually any visual content from a text description. The field transformed in 2022 with Stable Diffusion's open-source release and has continued advancing rapidly — FLUX.1 (Black Forest Labs, 2024) introduced flow-matching architecture that produces sharper text rendering and more accurate prompt adherence than standard DDPM diffusion. These models are now embedded in creative professional workflows, marketing production pipelines, and developer product features worldwide.

Image Generation Capabilities

  • Text-to-image — generate any visual from a natural-language description, positive and negative prompts
  • Inpainting — replace a selected region of an existing image while preserving the surrounding context
  • Outpainting — extend an image beyond its borders in any direction maintaining visual consistency
  • ControlNet / depth conditioning — constrain composition by pose, edge map, depth, or reference image
  • Style transfer — apply the visual style of a reference image to new generated content
  • High-res upscaling — add fine detail and resolution to low-res outputs using super-resolution models

Best Image Generation Models in 2026

FLUX.1 [dev] and FLUX.1 [schnell] (Black Forest Labs) set the current quality bar for open-weight models — exceptional text rendering and photorealism, Apache 2.0 licensed for schnell. Stable Diffusion 3 Medium is the community workhorse, supported across every major inference platform. DALL-E 3 (OpenAI) leads proprietary image APIs for prompt accuracy and safety. Ideogram V3 specialises in graphic design applications requiring accurate text-in-image rendering, achieving 90–95% text accuracy. Adobe Firefly is the choice for teams needing commercially safe content with full IP indemnification.