MovieRoll by Tencent
makes whole minute-long clips without forgetting who is in the scene.
- feed it a new prompt every ten seconds as the story changes
- no identity amnesia
- smart memory, gpu does not crash
- streams video in real
tencent-hunyuan.github.io/Mo…
π4 - 0.5B foundation model for real-time 3D reconstruction and camera pose estimation from uncalibrated RGB images.
- All-in-one in a single run
- No drifting tracks
- Handles massive scenes
cool for consumer drone and action camera mapping
yyfz.github.io/pi4/
MPEG-Diffusion
speeds up AI video models by up to 5x by rendering only the main subject in high res, keeping the bg low res
- Wan2.1, FLUX.2
- No visual glitches
- Lowers hardware memory and compute bottlenecks
dennismenn.github.io/mpeg-di…
Become a YouTube/TikTok creator with zero camera gear...
LongCat-Video-Avatar 1.6 just dropped. It's insane!
- No weird robot mouth
- 10min vids, no jitter
-Crazy fast
- Beats paid tools
streamers can sleep while the model talks
meigen-ai.github.io/LongCat-…
a handy tool if you need to protect vids/imgs in ComfyUI.
invisible watermarks, signed C2PA provenance and a verification receipt.
github.com/ariadne-mark/Comf…
LemongrassisLora
web studio for training LoRA/LoKr for Anima, Krea 2, MOSS Image.
- tag editor; projects, presets, honest VRAM/time estimates and crash-proof GPU queue
github.com/Lemongrassis/Lemo…
PAN 2 -General-purpose world model is now in ComfyUI!
- simulates action outcomes as video
-takes optional image+text observation and text actions
- LLM conceptual reasoning + 29B MMDiT
huggingface.co/Comfy-Org/PAN…
PAN 2 -General-purpose world model is now in ComfyUI!
- simulates action outcomes as video
-takes optional image+text observation and text actions
- LLM conceptual reasoning + 29B MMDiT
huggingface.co/Comfy-Org/PAN…
game dev is wild right now....
Qwen4-Max handles actual game engines natively
- give it one prompt and it builds scenes, assets, logic, and polish
With 71B active params, it should handle complex tasks pretty well. Need to test it in Blender, too
📢 Meet Qwen4-Max, your new AI teammate for work, creativity, and discovery!
Qwen4-Max is built on a new MoE architecture with native vision and a 1M-token context window.
Highlights:
- Outperforms Qwen3.8-Max with up to ~50% fewer tokens, while average cost per task drops over 60% in our internal evaluation.
- Advances real-world productivity through greater proficiency with professional tools, stronger long-horizon execution, and higher-quality deliverables.
- Handles real tasks, including 3D modeling, game development, manufacturing, finance, law, healthcare, and frontier scientific discovery.
- $1.5/M tokens for input, $4.5/M tokens for output, and $0.075/M tokens for implicit caching — all lower than Qwen3.8-Max, with implicit caching down by 70%.
The API is live on QwenCloud. You can also try Qwen4-Max on Qwen Studio, QwenWork and Qoder.
Start building today! 🚀
- Blog: qwen.ai/blog?id=qwen4-max
- API: qwencloud.com/models/qwen4-m…
- Token Plan: qwencloud.com/pricing/token-…
More to come! The open weights of smaller Qwen4 models will be released. Stay tuned. 👀
Claude + Blender? Nah, theres a new 3D beast in town-Ling-3.2!
"make a grim woods,rickety house, doodler is walking around", and it:
- Directs the entire clip
- Pulls all the assets
- Animates everything
- Renders out the final video
not quite perfect, but it looks solid!
Introducing Ling-3.2-flash, Ant Ling's native multimodal model: 560B total / 25B active parameters.
Stronger complex visual understanding, long-horizon multimodal agents, 3D understanding & interaction, and medical intelligence.
Outscores Opus 5 (max) in release evaluations:
- OmniDocBench1.5: 92.21 vs 90.56
- ZEROBench(Sub): 43.71 vs 39.52
- Vision2Web: 85.86 vs 80.20
MiniMax Image HDR VAE
keeps high-frequency details instead of smoothing them into mush.
Tested across color, sharpness, and gradient energy.
Good upgrade for image reconstruction workflows.
huggingface.co/Felldude/Mini…
ERNIE 5.5 Flash Preview just landed! 🔥
Built on a new architecture designed for efficient inference, the latest open-weight foundation model in the ERNIE family delivers stronger performance in coding, 3D modeling, knowledge work and content creation.
It keeps decoding fast while posting competitive scores on Terminal-Bench, AutomationBench, and other benchmarks.
Here's the rundown:
- 280B total parameters, 14B active per token
- 1M-token context window
- Generation speed above 200 tokens/sec
Model weights will be released by November 25.
Who does Blender 3D modeling better?
Quick test with the exact same prompt:
- GLM-5.3 Flash: not bad
- Ernie-5.5 Flash: fast, great tool use, but lacks vision.
- MiniMax-m3.1-Flash: super impressive. Flawless, one-shotted.
Ernie seriously needs multimodal support.