NVIDIA’s Gamma-World enables four-agent cooperation at 24 FPS
LangChain introduces Context Hub for managing AI agent contexts and teams up with MongoDB, Elastic, Pinecone, and Redis to create an open standard for agent memory.
NVIDIA
Most Recent
Jensen Huang Joins Tsinghua Board as Nvidia Cedes China Market to Huawei
Antigravity adds a leaner Gemini model to cut token waste on simple tasks and resets all account limits for the week.
Baseten revenue surges 20x, targets $11B valuation in $1B raise
Antigravity adds a leaner Gemini model to cut token waste on simple tasks and resets all account limits for the week.
AlphaGo’s search trick now generates longer, smoother AI videos
Antigravity adds a leaner Gemini model to cut token waste on simple tasks and resets all account limits for the week.
Playground AI founder warns of dual-track AI open source fracture
The White House secretly approved a $9 billion budget for NSA and CIA to buy Nvidia chips and deploy Anthropic's classified model Mythos for cybersecurity, bypassing Pentagon restrictions to avoid surveillance on US citizens.
AMD's MI355X challenges Nvidia B200 in long-context AI inference
Tencent's new Chronicles-OCR benchmark reveals that top AI models like GPT-5 and Gemini fail miserably at recognizing ancient Chinese scripts, scoring near zero on oracle bone and bronze inscriptions.
ADI to Acquire Empower in $1.5B AI Power Chip Deal
Tencent's new Chronicles-OCR benchmark reveals that top AI models like GPT-5 and Gemini fail miserably at recognizing ancient Chinese scripts, scoring near zero on oracle bone and bronze inscriptions.
Google and Blackstone bet $5 billion on a TPU rental startup to challenge Nvidia's
Tencent's new Chronicles-OCR benchmark reveals that top AI models like GPT-5 and Gemini fail miserably at recognizing ancient Chinese scripts, scoring near zero on oracle bone and bronze inscriptions.
Nous open-sources Lighthouse Attention, speeding up 512K context by 17x on a single B200
Elon Musk says a new 1.5T-parameter Grok model will launch within a month, and it's being trained on code from Cursor to fix data issues.
A 1GW AI data center will cost $38 billion
A six-person startup called Poetiq built an API-only harness that boosted Kimi K2.6 by 29.9 percentage points and let the lightweight Gemini 3.0 Flash beat bigger models like Claude Opus — all without fine-tuning any model weights.