Meta AI Released "Llama 3.2": Lightweight, Open, And Customizable models

Choose from 1B, 3B, 11B or 90B, or continue building with Llama 3.1

OpenAI Advanced Voice Mode🎤, NVIDIA & Llama-3.1-Nemotron-51B💻, Mercado Libre & Verdi🛒

Good news for ChatGPT Plus and Teams users, NVIDIA's Llama-3.1-Nemotron-51B model runs efficiently on a single H100 GPU, Mercado Libre uses the Verdi AI platform for efficient automation

NVIDIA’s Llama-3.1-Nemotron-51B: Redefining Efficiency in Language Models

Nemotron-51B-instruct is a model which offers a great tradeoff between model accuracy and efficiency

Mercado Libre Unveils Verdi: A GPT-4o Powered AI Developer Platform

OpenAI & Jony Ive Collaboration🤝, PUDU D7 "Semi-Humanoid Robot"🤖, OpenAI Academy🎓, Microsoft "GRIN-MoE"💻

Jony Ive and Sam Altman are working on an undisclosed AI computing project, Pudu unveils super semi-humanoid robot with 8-hour battery, OpenAI Academy launches with $1M in developer credits for devs in low- and middle-income countries.

OpenAI Academy: Empowering Global Developers with AI Tools and Resources

Microsoft Launched "GRIN-MoE" Gradient-Informed Mixture-of-Experts

Model for Efficient and Scalable Deep Learning: GRIN MoE with 6.6B activate parameters

Snapchat AI Video 👻, NVIDIA AI Aerial Platform 🛰️, 100 New Open-Source AI Models 🌐

From Alibaba’s Open-Source Models to Snapchat's and NVIDIA’s Game-Changing Innovations in Telecommunications and Social Media

Revolutionizing Telecommunications With AI-RAN and NVIDIA AI Aerial

Snapchat Unveiled "Snap AI Video"

New Snapchat Text-to-Video & Image-to-Video AI video-generation tool

Intel & AWS: Next-Gen AI Partnership 🤝, SambaNova’s Llama 3.1 Challenges OpenAI ⚡, Salesforce & NVIDIA: AI Agents 🦾

Intel and AWS are co-developing custom chips to revolutionize AI applications and U.S. manufacturing; SambaNova challenges OpenAI with record-breaking speed in its Llama 3.1 model; and Salesforce teams up with NVIDIA to launch advanced autonomous AI agents for enhanced enterprise productivity and real-time insights.

SambaNova Systems Challenges OpenAI with Llama 3.1-Powered Demo on Hugging Face

New speed record by processing 114 tokens per second