SliceOfAI

OmniHuman-1 📽️

An end-to-end multimodality-conditioned human video generation framework from Bytedance

🚀 OpenAI o3-mini: Faster, Smarter, and More Cost-Efficient Reasoning model

Yesterday, OpenAI launched o3-mini, their latest and most efficient model in their reasoning series, now available in ChatGPT & API!

🚀 Alibaba Launches Qwen2.5-VL: SOTA Vision model

OpenAI Operator 🤖 😱 : An Agent browsing and doing task on your behalf 😄

🤔 Some takes on: 1️⃣ Navigating Captcha, 2️⃣ Authentication for agents and, 3️⃣ Will agents have biases and it's societal implication!

Trump & U.S. Government Unveil Stargate: The Largest AI Infrastructure Project Ever! 🚀

A $500B Initiative to Secure America’s Leadership in AI

🚀 DeepSeek-R1: Advancing AI's Reasoning Capabilities and the Internet Can't Stop Talking! 🧠

DeepSeek has introduced the DeepSeek-R1 family, a series of models designed to push the boundaries of reasoning, math, and coding capabilities. With innovative training approaches, these models demonstrate strong performance across benchmarks like AIME 2024, MATH-500, and CodeForces.

Kokoro-82M, an open source, SOTA (State-of-the-art) TTS(Text-to-Speech) model 🗣️

A lightweight TTS model easy to setup, fast and open source for your projects.

Moondream 2B and 0.5B Vision model 👀

An open source lightweight vision model for your local AI stack