The AI & Creative Tools Magazine
New models, image and video generation, music generation, 3D generation, and the workflows creative people are building around them.
Our daily Hugging Face tracker flagged the ComfyUI repack of Mage-Flow climbing 35% in downloads overnight. The model behind it is small, fast, permissively licensed, and aimed squarely at people who generate images for a living.
Signal Board
Hugging Face 30→31 Jul 2026 · OpenRouter usage through 30 Jul 2026
36,180 downloads
Quantized repack of Kimi K3 — the format that decides whether a 2.8T model runs on your machine.
32,042 downloads
Weights for audio.cpp, a pure C++ engine running TTS, stems, and music generation with no Python.
60,162 downloads
ComfyUI repack of Microsoft's 4B native-resolution image model, MIT-licensed and merged into core.
664 likes · 0 downloads
Published this morning. Likes arrive first; downloads lag by a day or two — a launch caught mid-flight.
0.78T tokens · 10.1M requests
Fastest-growing model on the paid board among those already past a 50B-token base.
11.7% of image-model traffic
Google now supplies 54.1% of image-model traffic overall, up from 44.4% the week before.
Two public scoreboards tell you what creative-AI practitioners are really using, as opposed to what got announced. Hugging Face publishes downloads and likes for every public model and dataset — downloads approximate how much software is pulling a checkpoint, likes approximate how many humans thought it mattered. OpenRouter, which routes API traffic to hundreds of hosted models, publishes the tokens actually spent through it. We snapshot the top 500 of each daily and diff them, because the movement is the story: a launch is a press release, but a jump in downloads is somebody's pipeline changing.
The first full day of day-over-day data makes an unglamorous point. The three fastest-climbing repositories on the Hub were not new models at all — they were a GGUF quantization of Kimi K3 (up 197% in 24 hours), a ComfyUI repack of Microsoft's Mage-Flow (up 35%), and the GGUF bundle for audio.cpp (up 42%). None of those organizations trained anything. What they did was convert weights into a format that runs on hardware people own, inside software people already have open. That conversion layer is where a release stops being an announcement and starts being a tool, and it is consistently invisible in launch coverage. The counterweight sits at the top of the trending list: DeepSeek's V4 Flash, published this morning, has 664 likes and literally zero downloads — pure anticipation, with adoption still a day or two behind. Meanwhile on the paid side, Claude Opus 5's traffic through OpenRouter grew 30% in a day, and in image generation Google now supplies 54.1% of all traffic, with Gemini 3.5 Flash Lite taking 11.7% of the category in its first complete week.
The gap between 'that model exists' and 'that model runs on my GPU' just narrowed. Nunchaku's 4-bit inference is now native to Hugging Face Diffusers — no separate library, no local compilation.
Describe a game in a sentence; get a playable 3D world back. Roblox's Build tool just entered public alpha — mobile-native, AI-generated, and aimed at collapsing the distance between having an idea and shipping it.
Black Forest Labs made its name on FLUX image models. FLUX 3 is a different animal: one architecture that does images, video-with-audio, and robot actions — and, eventually, an open-weight release.
Texturing is one of the slowest steps in 3D. Abstract's InstaMAT update pairs generative materials with a full 2D painting toolset — the bet being that artists want AI to speed the work up, not take the brush away.
The fight over AI music is really a fight over training data. BandLab just bought Aiode — an AI studio whose whole pitch is that 100% of its training audio is licensed and traceable to the musicians who played it.
The interesting move in AI-for-art right now isn't another image generator — it's agents that reach into the tools artists already use. NVIDIA is wiring them into Blender, Houdini, Unreal, and Adobe through an open protocol.
Meta's first in-house image model, Muse Image, can blend several of your photos into one polished result and drop it straight into a chat or story. The launch was fast; the backlash over training data was faster.
Type a prompt or drop in a photo, and get back a 3D world you can move through — walls that stay put, depth you can step into. World Labs is making scene generation a creator tool, and it changes what 'generate an image' can mean.
Most generative AI tools produce an asset. Adobe's creative agent is explicitly built to leave the asset-making to specialized tools it already owns, and instead handle the orchestration between them.
Every long AI video you've seen until now was clips stitched together, seams and all. Seedance 2.5 generates the full 30 seconds in a single pass — and takes 50 reference inputs while doing it.
Sora launched to enormous hype and a viral app-store run. Eighteen months later, OpenAI is shutting the whole thing down — and Disney found out its billion-dollar partnership was dead less than an hour before the rest of us did.
Not every 'open' model is open the same way. FLUX.2's three-tier release is a good case study in reading the license before you build on top of any of them.
Most open video models still leave sound as someone else's problem. LTX-2.3 doesn't, and that one decision is why it's become a reference point for the rest of the field.