Models — page 5
Naveen Rao's Unconventional AI claims a 1,000x cut in inference power
The ex-Databricks AI chief released Un-0, an image model running on a simulated oscillator chip, with silicon schematics promised next.
IBM unveils nanostack architecture, claims first sub-1nm chip tech
The 0.7nm node packs nearly 100 billion transistors on a fingernail-sized chip, with 50% more performance than 2nm silicon.
Google bakes computer use into Gemini 3.5 Flash as a native tool
DeepMind folds its standalone agent model into Flash, letting developers build agents that drive browsers, mobile apps and desktops via one API call.
OpenAI unveils Jalapeño, its first custom inference chip built with Broadcom
The processor targets inference workloads and was co-developed using OpenAI's own models, with early tests showing better performance-per-watt.
GPT-5 Pro helps immunologist crack a 3-year T cell mystery
OpenAI says its top reasoning model gave Derya Unutmaz the breakthrough insight that closed a stalled immunology project with cancer implications.
Nvidia silicon runs 400 of the world's 500 fastest supercomputers
Nvidia GPUs, CPUs and networking now power 81% of the TOP500 and 90% of new entrants, with a clean sweep of the Green500 top eight.
Nvidia's Vera CPU lands at Los Alamos in three new supercomputers
Mission, Vision and Veritas will pair Vera CPUs with Rubin GPUs to run agentic science workloads, posting 7x gains over the Crossroads x86 system.
NVIDIA's new CUDA-X science software posts 14,900x speedup on telescope data
cuPhoton, DAQIRI, and ALCHEMI accelerate astronomy, particle physics, and materials simulation, with Lila Sciences cutting materials runs from weeks to days.
NAIRR pilot tops 700 research projects on NVIDIA DGX infrastructure
The NSF program's two-year run has accelerated fusion materials, infectious-disease detection, and fluid simulation foundation models.
JUPITER, Europe's first exascale supercomputer, posts results across brain, climate, and quantum
Four projects on the Jülich machine show what 20,480 Grace Hopper Superchips can do — from 1-km climate runs to a 50-qubit simulation.
Nvidia's Rubin AI servers run on 45°C coolant, killing data-center water use
The first 100% liquid-cooled platform from Nvidia targets $4M in annual savings per 50MW and a 100% cut in cooling water.
Subquadratic's SubQ posts third-party benchmarks for its sparse-attention LLM
The Miami startup says SubQ ran Nvidia's RULER 128 test for $8 versus $2,600 for Claude Opus 4.6, and Appen's evaluation backs the claim.
Nvidia's ENPIRE lets AI coding agents train robots overnight, hits 99% success
Teams of up to 8 agents from OpenAI, Anthropic and Moonshot taught robots to insert GPUs and cut zip ties without human input.
Nvidia Blackwell sweeps MLPerf Training 6.0 across all seven benchmarks
GB300 NVL72 delivers 1.6x faster training than GB200 NVL72, and CoreWeave hits a 2.02-minute time-to-train on DeepSeek-V3 671B.
Loft Orbital's Yam-9 runs first vision-language model in orbit
A Google DeepMind Gemma 3 model running on a Nvidia Jetson chip aboard Yam-9 identified targets from natural language queries.
NVIDIA Blackwell runs 20x more agents per megawatt on new AgentPerf benchmark
Artificial Analysis launches the first agentic AI benchmark, and the GB300 NVL72 takes the top spot against Hopper on DeepSeek V4 Pro.
Avataar's Varya undercuts Veo and Runway on price by 20x for Indian video AI
The Peak XV-backed startup distilled Alibaba's Wan 2.2 into a four-step model that generates 720p clips for $0.005 per second.
Anthropic reverses hidden Claude Fable 5 sabotage of AI researchers
After backlash, Anthropic will make its frontier-LLM safeguards visible instead of silently degrading Claude's output for rival AI work.
Anthropic's Claude Fable 5 refuses basic biology questions by design
Anthropic told The Verge Fable's guardrails are 'overly conservative' to block bioweapons queries, routing routine biology asks to Opus 4.8.
Writer researchers find memory tools make AI models more sycophantic, less accurate
Two new papers show that storing user preferences pulls models toward wrong answers, with Mem0 and Zep amplifying the bias.
Google DeepMind's DiffusionGemma generates text 4x faster on GPUs
The 26B MoE model hits 1,000+ tokens per second on an H100 by drafting 256 tokens in parallel instead of one at a time.
Decart launches Oasis 3, a real-time world model for driving simulation
The $4B startup is pricing API access at $0.02 per second and betting developers will build on top of it like the early OpenAI ecosystem.
Anthropic's Claude Fable 5 spins up playable games from a single prompt
Wharton's Ethan Mollick says the new Mythos-class model executed multi-page specs autonomously for up to a dozen hours.
Anthropic releases Claude Fable 5, its first public Mythos-class model
The lab once said Mythos was too dangerous to ship. New safeguards route 5% of risky prompts back to Opus 4.8.
Stay ahead of everyone in AI.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at