All updates

Got a low VRAM card? Or perhaps you just like editing images REALLY quickly? Then why not give Mage-Flow a try? Mage-Flow is a compact 4B-scale stack from Microsoft Research for speedy text-to-image generation, as well as instruction-based image editing. The stack is built from…

Two new INT8 models models are now supported by ComfyUI - JoyAI Image Edit and JoyAI Image Edit plus... but how do they fare again existing models such as Qwen Image Edit, Krea-2 and Flux.2 Klein? Image Edit focuses on editing of just a single image, whereas Image Edit Plus can…

Hi everyone! We are launching our V8.2 image model today. This update focuses on aesthetics and image quality and personalization. Images should be more creative, bold, sophisticated, edgy and fresh Random situations where you'd get a low quality image should be dramatically…

As a kid, all I wanted was a cool-place to work on cool-projects with cool-people. Projects to help people and push the world towards beauty, awe, and splendor. As an adult I find myself in the unusual position of having everything I wanted and two questions

Say goodbye to FP8 models and hello to INT8 Conv-Rot. Faster with more clear details? Yes please! In this video I compare some of the latest models to have conversions to INT8 including Qwen Image Edit 2511, Boogu Image Edit, Z-Image-Turbo, Krea-2 Turbo and SeedVR2 using my…

Krea-2 Turbo can be loads of fun, and if you like being creative then why not play with composite images and character creation? Building on my previous workflow tutorials, in this video you'll learn about 2 method for compositing, and also a 1-click random character creator…

At @fal we serve Ideogram V4 at high speed. At 1K resolution, a generation that used to take 2.75 s now renders in 0.44 s, a 6x speedup with no visible quality loss. We got there by pushing the cost of a generated image down on every axis

At @fal , we've achieved 16 times higher throughput on Qwen3.6 for a use-case that required high interactivity per user using DSpark on SGLang. But why did we need it? And how did we achieve it? Intro Text-to-image models like Ideogram V4 can generate stunning,

Ever wanted to use more than 3 reference images in your Krea-2 ComfyUI workflows? Perhaps you'd just like MORE ways than using the Qwen Image Edit Plus node for your references? No worries - got you covered. I've been super busy testing all sorts of new nodes for your Krea-2…

Heya everyone! We're making it easier to explore different styles in draft mode for V8.1. Include --sref random in your draft-mode prompt to create 24 images with different styles To turn on draft mode, click the ⚡ icon in the prompt bar or include --draft in

Krea-2 is now supported in ComfyUI with a wide range of model options available to run at home, on your own computer. The tiny, fast NVFP4 model is great for most people, with larger models available all the way up to the huge BF16. Their Turbo model needs just 8 steps, meaning…

In a world of slop, creatives need an aesthetics-focused model. Here's how to work with it.

Bernini-R now has native support in ComfyUI and provides a great way for you to do both image and video editing. Bernini-R uses the Wan2.2 architecture, employing two separate transformer models for high and low noise denoising, and so the existing lightx2v LoRAs work well to…

We'll be unveiling our first secret hardware project tomorrow, Wednesday 6/17. It's ambitious, (physically!) big, and unexpected. There will be a livestream at 6pm Pacific time both on Discord and X Discord Event For X, follow us on Twitter

There are so many FOSS models with "commercial use allowed" licenses out there now... but it can be difficult to know which one to use for the task at hand. Or perhaps you do know and you'd simply like to easily select your favourite model for your specific refining task…

By Emir Lise, Technical Partnerships Manager, AI Infrastructure @fal Today fal announced a strategic partnership with Amazon Web Services (AWS). I interned at AWS in college, working on compute pricing. What surprised me most was how many people were in the room for any given…

If you have never tried a video model before, now is the time.

AI image models have become extremely good at producing something that looks like stone, ceramic, brushed metal, leather, or concrete, but these images are not readily usable in a traditional CGI workflow - they have baked-in specular highlights, occlusion, perspective…

Starting today, HeyGen's models are available on fal. Whether you need talking avatars, automated video generation, or multilingual lip sync, HeyGen brings a focused set of capabilities that complement what's already on the platform. What HeyGen Brings Video Agent API — Give it…

Seedream 5.0 Lite dropped in February 2026 and it's a different beast from what we've been used to. Every single image on this page was generated with the exact prompt shown. This is an independent guide based on our own testing. We just really like

Seedream 5.0 brings multi-step reasoning, example-based editing, and deep domain knowledge to image generation. Here's what you should know.

Recraft V4 generates art-directed images — and actual editable SVGs — with strong composition, accurate text rendering, and what the Recraft team calls "design taste." Four models are available on Replicate now.