LCLMs compress LLM context before decode — 8.8x faster at 16x compression, beating every KV cache method tested. Open-sourced by NYU and Columbia.

Microsoft's SkillOpt brings deep-learning discipline to AI agent skills, replacing manual prompt tweaking with mathematically ...

Enterprise AI teams have spent years solving for compute, securing GPU allocations, negotiating cloud capacity, and ...

Google's open-source diffusion language model generates 256 tokens in parallel and self-corrects, hitting 4x speed on one GPU ...

As AI continues to evolve, leaders should invest not only in tools, but also in R&D processes and cultural foundations that ...

Sapient researchers trained a 1B reasoning model on just 40B tokens — scoring competitively with 2B-7B models at a fraction ...

The victory of GPT-5.5 aligns with recent third-party analysis suggesting that OpenAI's models are currently superior at ...

MassMutual caps vendor contracts at 12 months and runs a multi-model architecture — cutting contact center resolution times ...

Collaboration will enable better decisions, higher treatment quality, and scalable clinical impact for better outcomes for patients MUNICH & REDWOOD ...

RAAPID's AI-powered risk adjustment solution received an A+ would-buy-again grade (Emerging Data, n=5) from interviewed customers in the May 2026 KLAS ...

Aryon Security, a Cloud Security Enforcement Platform, today announced it has raised $29 million in Series A funding led by ...

Cohere's North Mini Code ranks 8th of 127 open-weight models on output speed — but generates 3x the output tokens of ...