2 results found
NVIDIA Nemotron 3.5 Lightning is an open-weights, hybrid MoE (Mamba-2 + MoE + Attention) LLM designed for efficient AI agent development. It supports 1M token context, offers speculative decoding methods like DSpark, and is optimized for NVIDIA Blackwell, Hopper, and Ampere GPUs. Developers can deploy it via vLLM, TensorRT-LLM, or SGLang, leveraging its advanced features like reasoning control and tool-calling through an OpenAI-compatible API.
Microsoft has launched Phi-4-reasoning-vision-15B, a compact multimodal AI that intelligently decides when to apply complex reasoning and when to respond directly. This open-weight model matches larger systems' performance with significantly less data, signaling a shift toward efficient, practical AI deployment across various applications.
