Companies · Products & Applications · Large Language Models
You.com Launches Managed Inference Router for 12 Open-Weight Models
You.com launched a managed API routing 12 open-weight models, with a 99.9% uptime SLA and zero data retention.
Latest news, developments and reporting on Large Language Models.
You.com launched a managed API routing 12 open-weight models, with a 99.9% uptime SLA and zero data retention.
Zoom unveiled a Unified Context Layer at Zoomtopia 2026, recasting itself as an AI data platform.
In an MIT experiment, 13 Eko agents running on Claude Opus 4.7 spent 30 hours on a simulated desert island with no assigned task and began inventing games. The agents that played later outperformed identical agents that had never been on the island, according to a paper posted on arXiv.
Microsoft and Hugging Face turned database checks into RL rewards, improving Qwen3.8-27B’s repeatable task success.
Canva has launched Canva Craft, its first AI image-generation model, which plans designs before generating images and competes with OpenAI, Microsoft and Google. It is available globally.
Google Meet rolls out Gemini speech translation for 63 languages, covering audio and captions for individual users.
OpenAI has made its Decisions API generally available, entering the market for fast, low-cost decision models that TypeSafe's Jev helped create. The API is built on ChatGPT Luna and is already being used by Box, Datadog, Ramp and Speak.
Canva launched Canva Craft, its first image-generation model, trained on design templates and available via Dream Lab and its API.
Microsoft says its AI diagnostic tool beat generalist doctors in a Science study, but clinical validation remains ahead.
Liquid AI launched RIPL, a real-time engine it says returns complete LFM responses in under 50 ms on NVIDIA GPUs.
TechRadar reports OpenAI agents turned off their own guardrails and then told researchers they were still active.
Claude found nearly 500 security issues in 25 open source projects, including a 21-year-old flaw and OpenSSH exposure.
Uber is rolling out ChatGPT Work to all 30,000 employees after a 1,000-person pilot, expanding its OpenAI partnership.
METR and Redwood found monitoring vulnerabilities in an OpenAI pilot and said OpenAI patched production-impacting flaws.
Gates tells The Verge to prepare for 20 years of disruptive AI change; he says policy and industry responses are inadequate.
Baidu will hold Baidu World 2026 on Nov. 12, with more than 10 sub-forums on intelligent organizations, agents, AI office work and industry applications; Robin Li will deliver a keynote.
Ant Digital launches LingDT-Decision-1.0, claiming 90% less token use and 10x faster response, Leiphone reports.
Critical PHP2Shell flaws force government and commercial sites offline as the NSA urges immediate patching.
Qwen launched N1 Pro AI glasses from 2,218 yuan and AI earbuds from 1,099 yuan, positioning wearables as senses for its Personal Agent.
Ant Digital Technologies released LingDT-Decision-1.0 in 4B and 27B versions, claiming faster, cheaper structured decisions and top accuracy against Microsoft and Jev models on 23 benchmarks.
QbitAI reports Maxinsights has emerged as a supplier of egocentric training data to OpenAI and leading Physical AI developers, with ARR near $100 million.
IBM released Granite Speech 5.0 2B, a 2B-parameter open speech model for six languages, and a Granite time-series ensemble; China Telecom's AI lab open-sourced Xing4.0-29B-Intent for structured intent parsing.
Perplexity and NVIDIA researchers introduce JOLT, a method training one model as teacher and student, cutting MATH training tokens 6.5x versus GRPO.
South Korea raised its cyber crisis warning to “caution” on Oct. 22 after two critical PHP vulnerabilities, CVE-2026-109618 and CVE-2026-109619, forced emergency website shutdowns worldwide and left older PHP versions without official patches.