Thanks Matthew! I really want to get this working super well. My next step is trying to get it to generate 2x speed videos… frame rate takes a hit but it may enable longer videos w/ fewer jarring transitions
SOFTWARE
-
Multiple Projects Development Staggered Release Approach
By
–
lots of people working on lots of things, were gonna release what we can when we can, it's not all zero sum
-
Great Web App Works Perfectly on Android Platform
By
–
there's a great web app and the Web app works great on Android too
-
Imagine v0.9 Native Audio Video Generation Creates Cinematic Experiences
By
–
Imagine v0.9 pushes the boundaries of native audio + video generation, creating cinematic experiences straight out of the box—no editing required.
— xAI (@xai) 7 octobre 2025
Watch this video of a dragon with synced, immersive sound. Generated by @GlitterPixely. pic.twitter.com/3Wh9YBIY0kImagine v0.9 pushes the boundaries of native audio + video generation, creating cinematic experiences straight out of the box—no editing required. Watch this video of a dragon with synced, immersive sound. Generated by @GlitterPixely
. -
Scaling TorchTitan with SkyPilot Community Tutorials
By
–
More ways to scale TorchTitan with SkyPilot great to see the community expanding the ecosystem with new tutorials like this one!
-
Sora 2 Pro Launches With Longer Complex Scenes Higher Resolution
By
–
Sora 2 Pro produces longer, more complex scenes at higher resolution than base Sora 2. Sora 2 Pro is available at https://
poe.com/Sora-2-Pro, while Sora 2 is at https://
poe.com/sora-2. You can try both across all platforms, and can be used through the Poe API. Enjoy! (2/2) -
Google Adds Annotations to AI Studio Build
By
–
Google is working on "Annotations" support for AI Studio Build. Users will be able to draw, leave comments and highlights on the Canvas so Gemini can follow more granular instructions.
— 🚨 AI News | TestingCatalog (@testingcatalog) 7 octobre 2025
PR Review >>> UI Review 👀 pic.twitter.com/VktKU0FNQRGoogle is working on "Annotations" support for AI Studio Build. Users will be able to draw, leave comments and highlights on the Canvas so Gemini can follow more granular instructions. PR Review >>> UI Review
-
Xet: Upload Only New File Chunks for Efficiency
By
–
Powered by Xet so you only upload the new file chunks, not the whole file
-
LFM2-8B-A1B: Efficient MoE Language Model Released
By
–
LFM2-8B-A1B just dropped on @huggingface!
— Maxime Labonne @ ICLR (@maximelabonne) 7 octobre 2025
8.3B params with only 1.5B active/token 🚀
> Quality ≈ 3–4B dense, yet faster than Qwen3-1.7B
> MoE designed to run on phones/laptops (llama.cpp / vLLM)
> Pre-trained on 12T tokens → strong math/code/IF pic.twitter.com/cGbuJoOMDNLFM2-8B-A1B just dropped on @huggingface
! 8.3B params with only 1.5B active/token > Quality ≈ 3–4B dense, yet faster than Qwen3-1.7B
> MoE designed to run on phones/laptops (llama.cpp / vLLM)
> Pre-trained on 12T tokens → strong math/code/IF -
ade-python: Transform Complex Documents into LLM-Ready Data
By
–
Turn complex and messy documents into LLM-ready data! ade-python is a Python library for Agentic Document Extraction (ADE) that outputs layout-aware structured JSON from visually complex documents.