🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Our vision is for AI that uses world models to adapt in new and dynamic environments and efficiently learn new skills. We’re sharing V-JEPA 2, a new world model with state-of-the-art performance in visual understanding and prediction. V-JEPA 2 is a 1.2 billion-parameter model,

Most of the older generation will limit themselves by overproducing their input. It will likely be the kids that really show us what it means when anything can be anything.

Seedance 1.0 Exploring the Boundaries of Video Generation Models

Testing vertical video generation with #Veo3

⭐The Year of Inference is here. Featherless is now an official inference provider on Hugging Face, unlocking 6,700+ LLMs for anyone to run, eval, and deploy instantly. It all starts with accessibility. From DeepSeek to Mistral, LLaMA to Qwen — powerful LLMs are one click

These arcade cabinets are powered by three.js Here are the 3 browser APIs that make this possible 👀

fyi: midjourney video is actually COOL AF! 🎸

In 3 days I've already spent $100. This is why the Claude Code Max plan is worth it.

There's a new AI video model topping the leaderboard: Seedance. It's #1 in the Artificial Analysis ranking for text-to-video and image-to-video. And after voting on 3k+ clips across dozens of models, I found the hype holds up. 5 things it's particularly good at (with examples) 👇

2) Prompt adherence The model can take complex prompts and adhere to them - in terms of both subject and camera movement, as well as details of the environment. Prompts here: "have the gamer play with increasing intensity as the camera pushes in and pans right, tracking around

👇👇👇 Bytedance Seed's Seedream 3.0 and Seedance 1.0 Lite is now available on fal, as APIs! These models rank very highly on Artificial Analysis 's arenas, so give them a try!

Trying out the new Hunyuan3D 2.1 on fal - the PBR material is really neat! To test this out I downloaded the output zip and loaded the model into the three.js editor, and attached the texture, metal, and rough maps manually (and added directional and point lights)

we need to talk about midjourney video interactions.. WTF IS THIS???

FWIW, Isaac Sim just OSed

I've spent almost $1,000 in veo3 api costs this month, AMA

📣 Humanoids are a hot topic right now, everyone’s talking about them. It’s no surprise that the biggest tech giants are jumping on the bandwagon, eager to be part of the hype. CB Insights even put together a slick graphic showing just how deep these companies are getting into

Why can AIs code for 1h but not 10h? A simple explanation: if there's a 10% chance of error per 10min step (say), the success rate is: 1h: 53% 4h: 8% 10h: 0.002% Toby Ord has tested this 'constant error rate' theory and shown it's a good fit for the data chance of

Lerobot community stats after the worldwide LeRobot hackathon which happened this weekend: - 10k+ Discord members - 100+ Github Contributors - 2M+ datasets downloads - 10k+ datasets uploaded to the hub, the equivalent of 260 days of recording Crazy how this open-source

We've made substantial progress on our action-conditioned video generation model, aka the "1X World Model", and we show that we can use it to evaluate robot policies instead of running experiments in the real world. Check it out!

🚨 Hugging Face + Groq Inc = Game Changer for Builders Groq is now integrated into the Hugging Face Playground and API, unlocking the speed to bring your most ambitious ideas to life — faster than ever. From smarter agents to real-time copilots, the future is yours to