Explore AI models for media generation

Explore curated collections or search across 300K+ supported AI models for image and video generation, including checkpoints, LoRAs checkpoints, ControlNets, VAEs, and more. Test models in the Playground or integrate them instantly with our API.

Image collections

Video collections

Audio collections

Explore 312,807 community models

Checkpoint

15.0K

HunyuanImage-3.0 Tencent’s HunyuanImage-3.0 is an open-source, 80B-parameter MoE model that unifies text and image generation via an autoregressive framework, enabling superior prompt alignment with world-knowledge reasoning, and sharp text rendering.

runware:108@22

Checkpoint

36.3K

Qwen-Image-Edit-Plus An enhanced version of Qwen-Image-Edit that supports multi-image editing, improved consistency in single-image tasks, and native ControlNet conditioning for fine-grained control.

Launch in Playground

Checkpoint

9.5K

Vidu Q1 Maintains exact reference fidelity from input images; supports multi‑reference merging for characters, scene, props with natural blending. Ideal for ads, stories, animation, and creative image synthesis.

bytedance:5@0

Checkpoint

24.0K

Seedream 4.0 Next-gen multimodal AI model for ultra-fast high-res generation (2K–4K) with batch/coherent outputs, natural-language editing, sequential outputs, and bilingual Chinese–English support.

Launch in Playground

google:4@1

Checkpoint

136.0K

Gemini Flash Image 2.5 Built for complex visual tasks, it handles multi-image generation with coherent character identity while drawing on strong world understanding to guide edits and completions.

Launch in Playground

runware:108@20

Checkpoint

47.8K

Qwen‑Image‑Edit Designed for intelligent image editing, it applies targeted changes while keeping the original composition intact. Reliable results across a wide range of visual styles and prompts.

Launch in Playground

runware:108@21

Checkpoint

30.0K

Qwen‑Image‑Edit Lightning (8 steps) Fast 8-step version of Qwen‑Image‑Edit, combining high-speed inference with precise, localized image edits. Ideal for quick iterations without sacrificing control or consistency.

Launch in Playground

Checkpoint

18.1K

Kolors 2.1 Polished version of 2.0 with refined edge rendering and better lighting realism. Excellent general-purpose model for both creative and commercial use.

Checkpoint

12.7K

Kolors 2.0 A major leap forward with Kling 2.0 tech, delivering photorealism, natural color balance, and advanced compositional depth. Suitable for cinematic shots and human subjects.

Checkpoint

22.9K

BRIA 3.2 Commercial-ready text-to-image model with improved aesthetics, strong prompt alignment, and superior short-text rendering, trained on licensed data.

Checkpoint

14.0K

FLUX.1 [dev] SRPO 12B‑parameter flow transformer finetuned with SRPO on FLUX.1 [dev]. Produces highly realistic images with enhanced lighting, texture, and aesthetics while reducing the typical AI artifacts.

runware:108@1

Checkpoint

5.6K

Qwen-Image Qwen-Image is a vision-language model by Alibaba Cloud that can understand and generate images with strong detail and accurate prompt following.

Launch in Playground

runware:108@5

LoRA

12.0K

Qwen‑Image-Lightning (4 steps) LoRA-styled distillation of Qwen‑Image for 4-step inference, up to 25× faster with minimal visual quality loss.

Launch in Playground

runware:108@6

LoRA

13.7K

Qwen‑Image-Lightning (8 steps V1.0) 8-step distillation of Qwen‑Image offering a solid speed-quality tradeoff—faster inference with strong text and visual fidelity.

Launch in Playground

runware:108@7

LoRA

29.6K

Qwen‑Image-Lightning (8 steps V1.1) Improved 8-step distillation of Qwen‑Image with refined quality and consistent rendering at much faster speeds.

Launch in Playground

runware:107@1

Checkpoint

27.3K

FLUX.1 Krea [dev] FLUX.1 Krea [dev] is a photorealistic open-weight model co‑developed with Krea AI. It offers distinctive, realistic results without the usual AI feel, and serves as a strong base for custom generations.

Launch in Playground

bytedance:4@1

Checkpoint

155.5K

SeedEdit 3.0 SeedEdit 3.0 is ByteDance's advanced image editing model for precise, high-resolution control. It preserves unedited regions and follows prompts closely. Supports 4K output and generates results in under 15 seconds.

Launch in Playground

bytedance:3@1

Checkpoint

118.5K

Seedream 3.0 Seedream 3.0 is a bilingual (Chinese‑English) text‑to‑image model delivering fast, photorealistic images up to 2048×2048 resolution with industry‑leading text rendering and layout accuracy.

Launch in Playground

ideogram:4@1

Checkpoint

26.3K

Ideogram 3.0 Ideogram 3.0 pushes design-level generation to new heights, with sharper text rendering and better composition. It also adds greater stylistic control, perfect for graphic-driven content.

Launch in Playground

ideogram:4@2

Checkpoint

25.2K

Ideogram 3.0 Remix Enables reinterpretation of existing designs with fresh styles or palettes, while preserving structural intent. Useful for A/B design testing or creative variation.

Launch in Playground

ideogram:4@3

Checkpoint

26.8K

Ideogram 3.0 Edit An inpainting model that lets you surgically edit or replace parts of an image, great for polishing visuals or fixing text without starting from scratch.

Launch in Playground

ideogram:4@4

Checkpoint

21.0K

Ideogram 3.0 Reframe Expands visuals beyond their original borders using style-consistent outpainting, perfect for adapting content to new formats or aspect ratios.

Launch in Playground

ideogram:4@5

Checkpoint

19.2K

Ideogram 3.0 Replace Background Swap out the background while keeping foreground elements intact. Useful for product mockups or design overlays.

Launch in Playground

Checkpoint

12.2K

Flex.1‑alpha Compact 8B image model with optional guidance embedder, CFG-free generation, and Apache 2.0 licensing.

IP-Adapter

18.8K

Flex.1‑alpha Redux Adapter for Flex.1‑alpha that enhances spatial precision using a stronger vision encoder. Ideal for highly detailed and accurate image generation.

google:2@1

Checkpoint

101.5K

Imagen 4 Preview Preview version of Google's latest image generation model, offering 2K resolution, superior detail rendering, and improved text generation capabilities across various styles.

Launch in Playground

google:2@2

Checkpoint

185.9K

Imagen 4 Ultra Google's highest-quality image model to date, combining photorealism, sharp detail, improved text rendering, and fast generation in a single architecture.

Launch in Playground

google:2@3

Checkpoint

148.7K

Imagen 4 Fast A faster variant of Imagen 4 optimized for latency-sensitive use cases, delivering high visual quality with significantly reduced generation time.

Launch in Playground

openai:1@1

Checkpoint

23.5K

GPT Image 1 GPT Image 1 is OpenAI's latest image model (March 2025), built on GPT‑4o. It generates high‑quality, diverse visuals with advanced prompt following, text layout, and fast editing capabilities.

Launch in Playground

runware:106@1

Checkpoint

24.2K

FLUX.1 Kontext [dev] Open-weights iterative editing with local and full-scene changes, offering balanced quality and speed for development workflows.

Launch in Playground

bfl:3@1

Checkpoint

343.0K

FLUX.1 Kontext [pro] Fast, iterative editing with local + full-scene changes, preserving style across multiple.

Launch in Playground

bfl:4@1

Checkpoint

255.2K

FLUX.1 Kontext [max] Best quality and prompt accuracy with faster, sharper edits + premium typography support.

Launch in Playground

bfl:2@1

Checkpoint

170.4K

FLUX.1.1 [pro] Enhanced version with improved image quality, prompt adherence, and output diversity, delivering faster and more accurate results.

Launch in Playground

bfl:2@2

Checkpoint

141.4K

FLUX.1.1 [pro] Ultra High-resolution model capable of generating images up to 4 megapixels, offering ultra and raw modes for detailed and realistic outputs.

Launch in Playground

runware:97@3

Checkpoint

718.9K

HiDream-I1 Fast Built for speed. Great for fast drafts, real-time ideas, and bulk jobs. Surprisingly strong prompt adherence and structure for its size.

Launch in Playground

runware:97@2

Checkpoint

596.9K

HiDream-I1 Dev Balances quality and speed. Supports LoRAs and produces clean, detailed images. Ideal for testing ideas and refining styles.

Launch in Playground

runware:97@1

Checkpoint

306.5K

HiDream-I1 Full The highest quality HiDream model. Sharp detail, accurate prompts, and full LoRA compatibility. Fully uncensored for creative freedom.

Launch in Playground

rundiffusion:130@100

Checkpoint

332.7K

Juggernaut Pro Flux by RunDiffusion Combines Juggernaut Base with RunDiffusion Photo and features enhancements like reduced background blur.

Launch in Playground

rundiffusion:110@101

Checkpoint

432.7K

Juggernaut Lightning Flux by RunDiffusion Blazing-fast, high-quality images rendered at five times the speed of Flux. Perfect for mood boards and mass ideation, this model scales to both solo use and production.

Launch in Playground

Test and run inference instantly

Live Playground testing

Experiment with parameters and settings in our interactive Playground before deployment.

Simple API integration

Deploy models with straightforward API calls. No complex infrastructure setup required.

Global high-performance hosting

Models distributed across our network for low-latency, scalable inference worldwide.

UI Interface of Playground

Pick from 312,807 models

Whether it’s base models, custom checkpoints, LoRAs, ControlNets or embeddings, the entire Stable Diffusion ecosystem runs quickly and flexibly through the Runware API.

Access the latest AI technologies, including ControlNet, InstantID, AnimateDiff, IP-Adapters, and more. See all supported Image Gen technologies.

A jar of cream surrounded by white flowers and green leaves

A colorful teddy bear on a pink background

A pink robot with headphones on a pink background

A little girl dressed as a superhero with a batman mask

A pair of colorful shoes on a pink and orange background

A woman with glowing lines on her face

A group of windmills sitting next to a body of water under a colorful sky

A woman walking up a set of stairs with a brain inside of her head

A person standing in front of an elephant in a room

A man standing next to a man dressed as the Hulk

A blue handbag sitting on top of a pile of rocks

A panda playing a bass guitar in front of a crowd

The joker in batman arkham origins

A man in a purple hoodie with a monkey face on it

A drawing of a pug dog with different expressions

A cartoon of a man with black hair and a striped shirt

A black and white illustration of a pug wearing a military uniform

A couple of houses sitting on top of a lush green hillside

A 3D rendering of a chain on a black background

A woman with her eyes closed and her hair blowing in the wind

A can of fruit juice sitting on top of a table

A person riding a motorcycle on a city street at night

A black and white drawing of a pug singing into a microphone

A woman holding a gun in a futuristic setting