Stable Diffusion online AI

Gemma 4 just got even faster! We're releasing Multi-Token Prediction (MTP) drafters that deliver up to a 3x speedup, without any degradation in output quality or reasoning logic.

Gemma 4 just got a massive speed-up with MTP drafters ⚡️ > speculative decoding (up to 3x tokens/sec improvement compared to normal Gemma-4 🔥) > identical reasoning, just faster > day-0 support in transformers, MLX, vLLM > A2.0 licensed 🤗

Meet a powerful uncensored text model that's turning heads. This 4B parameter Flux2-Klein variant offers unrestricted text generation, perfect for developers who need creative freedom without guardrails. It's GGUF-quantized for efficiency.

ComfyUI Action Director. Interactive 3D viewport for ControlNet; - loads FBX/GLB; batch-renders OpenPose, Depth, Canny, Normal; - manual near/far depth control and infinite UI scaling. github.com/yedp123/ComfyU…

Kling 3.0 is now available in ComfyUI via Partner Nodes! You can now generate multiple shots in a single run with precise duration control for each. A new level of consistency, multilingual dialogue, and up to 15 seconds of cinematic output. Fully composable with ComfyUI’s ecosystem of nodes and extensions!

people sleep on last week's open multimodal releases > GLM-OCR: sota OCR model > MiniCPM-o-4.5: Gemini 2.5-flash level Omni model that runs on your phone > InternS1: efficient generalist VLM outperforming on science tasks all allow commercial use freely 🔥

I didn’t build another AI course. I built a world. Cartoon Hero World 🏰 A place to learn, practice, and turn AI skills into real opportunities. If you’re serious about AI animation in 2026 - this is where you start. Gates are open 👇

Today, we’re excited to share The Complete AI Upscaling Handbook. Upscale has been a big topic in the space. And after weeks of testing and integration, creators can now access nearly every major upscaling approach in ComfyUI. This is a deep dive into benchmarks, 10 real-world use cases, and 20 production workflows, to help you choose the best practice. 🧵 Try the attached workflows!

🚀 Introducing Qwen3-Coder-Next, an open-weight LM built for coding agents & local development. What’s new: 🤖 Scaling agentic training: 800K verifiable tasks + executable envs 📈 Efficiency–Performance Tradeoff: achieves strong results on SWE-Bench Pro with 80B total params and 3B active ✨ Supports OpenClaw, Qwen Code, Claude Code, web dev, browser use, Cline, etc 🤗 Hugging Face: huggingface.co/collections/Qw… 🤖 ModelScope: modelscope.cn/collections/Qw… 📝 Blog: qwen.ai/blog?id=qwen3-… 📄 Tech report: github.com/QwenLM/Qwen3-C…

Introducing Voxtral Transcribe 2, next-gen speech-to-text models by @MistralAI. State-of-the-art transcription, speaker diarization, sub-200ms real-time latency. Details in 🧵

Meet WAN2.2-14B-Rapid-AllInOne: an image-to-video powerhouse that turns static pictures into dynamic scenes. This model is blowing up because it transforms single images into full video sequences, opening new creative possibilities for AI video generation.

Imagine you want to create not just 1 short cartoon 🎬 You want to create a series. Same characters, different environments. This is how to get started👇

🚨 Calling all ComfyUI users in VFX & games We’re talking with teams to understand how real production workflows are running and use that input to improve the product. If your studio is open to a short conversation to help guide what we build next: 🔗 docs.google.com/forms/d/e/1FAI…

TII just dropped Falcon H1R-7B 🙌🏻 a new reasoning model outperforming others in math and coding with only 7B params with 256k context window 🔥 it's a mamba-transformers hybrid model so more efficient per throughput and memory too 🤩

AMD ROCm™ 7.1.1 builds on the foundation of ROCm 6.4 and delivers up to 5.4x performance uplifts in ComfyUI on Windows.

Most people still think AI can't do this. Meanwhile, creators are dropping fire anime fights and competing in tournaments. This way they’re not just learning AI animation and storytelling at lightning speed - they’re building their first IPs. Just AI + your vision.

Get started (Local ComfyUI): 1. Use the Qwen Image workflow in Template Library or open this Qwen Image 2512 workflow: github.com/Comfy-Org/work… 2. Download the models: huggingface.co/Comfy-Org/Qwen… 3. Edit your prompt and run!

Happy New Year! Looking forward to an incredible 2026 in the AI community. So thankful for all my friends here

Run Qwen-Image-2512 in ComfyUI: no update needed! This upgraded version of the Qwen Image text-to-image model brings: More realistic human rendering, with far less “AI look” Finer natural details across landscapes, fur, and textures Improved text rendering and layout accuracy ▶️ Run it locally in ComfyUI with the Qwen Image workflow ☁️ Comfy Cloud support coming soon

Make sure to update Grok every day. I notice improvement of the model between major releases.