OpenAI launches GPT-5.4 mini and nano for faster, cheaper AI
OpenAI is shrinking GPT-5.4 down into mini and nano models that stay fast, cheap, and surprisingly close to flagship performance.
Perplexity unveils Comet Enterprise with granular admin and security controls
The new Comet Enterprise offering lets admins decide exactly where AI can operate, from specific domains to full-browser access, bringing genuine governance to AI-native…
Perplexity enters NVIDIA Nemotron Coalition as a founding partner
As a founding member of NVIDIA’s Nemotron Coalition, Perplexity is helping turn frontier‑grade base models into open infrastructure for developers…
Google DeepMind maps a new way to score AI systems on the road to AGI
Google DeepMind is done with vague AGI talk and is rolling out a cognitive scorecard that measures what today’s models…
Special-edition Nike Powerbeats Pro 2 land with Volt design and ANC
With dual branding, a Volt-splatter case and secure-fit earhooks, the Nike Powerbeats Pro 2 are tuned for people who live…
YouTube is now a preferred platform for the FIFA World Cup 2026
For the first time ever, fans can catch live kickoffs,…
Google’s Colab MCP server lets any AI agent run your notebooks
Google’s Colab MCP server uses the Model Context Protocol so…
Amazon launches ultra-fast 1-hour and 3-hour delivery in more US cities
The company is layering 1-hour and 3-hour delivery on top…
WhatsApp tests noise cancellation for Android voice and video calls
WhatsApp is catching up to system‑level audio tricks by building…
You can now use WhatsApp on select Garmin smartwatches
With native WhatsApp support, compatible Garmin watches now show recent…
Apple Developer lands on bilibili and LinkedIn
Developers can now follow Apple Developer on bilibili and LinkedIn…
NVIDIA Vera Rubin POD unites seven chips into one AI powerhouse
Rather than overprovisioning power, the MGX-based Vera Rubin POD dynamically…
Local‑first OpenClaw agents on RTX and DGX Spark
NVIDIA’s RTX GPUs and DGX Spark give OpenClaw the compute…