Xiaomi stuns with new MiMo-V2-Pro LLM nearing GPT-5.2, Opus 4.6 performance at a fraction of the cost

MiMo-V2-Pro utilizes a 7:1 hybrid ratio (increased from 5:1 in the Flash version) to manage its massive 1M-token context window.

New MiniMax M2.7 proprietary AI model is 'self-evolving' and can perform 30-50% of reinforcement learning research workflow

For direct API integration and via third-party provider OpenRouter, MiniMax M2.7 maintains a cost-leading price point of 0.30 dollars per 1 million input tokens and 1.20 dollars per 1 million output ...

Hirundo Uses NVIDIA NeMo Evaluator, CUDA, and GB200 NVL72 to Validate Breakthrough AI Safety Results Across Open-Source LLMs

Machine Unlearning platform powered by the NVIDIA stack demonstrates up to 91% reduction in prompt injections and 95% reduction in bias across foundat ...

RAAPID Secures Series-A Extension to Scale Neuro-Symbolic AI-Powered Medical Coding

UPMC Enterprises joins M12 (Microsoft's venture fund) in backing RAAPID's compliance-first Clinical AI Platform for Medicare Advantage risk adjustment ...

Enterprise AI agents keep operating from different versions of reality — Microsoft says Fabric IQ is the fix

Microsoft has made Fabric IQ's business ontology accessible via MCP to any AI agent from any vendor, targeting the core reason multi-agent systems fail in production: agents operating from conflicting ...

Astro Pak LLC Completes Acquisition of Clean Sciences

Astro Pak LLC (“Astro Pak”), a portfolio company of The Stephens Group, LLC (“Stephens Group”), is pleased to announce its acquisition of Clean Sciences, LLC (“Clean Sciences”), a precision cleaning ...

Nvidia shrinks LLM memory 20x without changing model weights

Nvidia's KV Cache Transform Coding (KVTC) compresses LLM key-value cache by 20x without model changes, cutting GPU memory costs and time-to-first-token by up to 8x for multi-turn AI applications.

Open source Mamba 3 arrives to surpass Transformer architecture with nearly 4% improved language modeling, reduced latency

This release is good for developers building long-context applications, real-time reasoning agents, or those seeking to reduce GPU costs in high-volume production environments.

The authorization problem that could break enterprise AI

When an AI agent needs to log into your CRM, pull records from your database, and send an email on your behalf, whose identity is it using? And what happens when no one knows the ...

Nvidia's agentic AI stack is the first major platform to ship with security at launch, but governance gaps remain

Five security vendors shipped governance for Nvidia's agentic AI stack at GTC — the first time security has launched with a major AI platform. Here's the five-layer framework, what it covers, and ...

Nvidia lets its 'claws' out: NemoClaw brings security, scale to the agent platform taking over AI

That said the direction is clear. Claws are coming to the enterprise. Nvidia just made its bet on being the platform they run ...

Langsmart Publishes Industry’s First p95 Semantic Cache Benchmarks for On-Premises AI Gateway, Challenges Market: “Show Me the p95”

Testing Confirms 10.2x Faster Response Times, Exceeding Cloud-Hosted Alternatives SAN JOSE, Calif.--(BUSINESS WIRE)--March 17, 2026-- NVIDIA GTC 202 ...