NVIDIA Releases Nemotron 3 Nano Omni Multimodal Agent Reasoning Model
NVIDIA has launched the Nemotron 3 Nano Omni open-source model, integrating vision, speech, text, and other multimodal r…
📂 LLM News
Latest articles in LLM News
Xiaomi Releases MiMo-v2.5 Series Model Weights
Xiaomi has officially open-sourced the MiMo-v2.5 series large language model weights, delivering strong performance in c…
SenseTime Open-Sources SenseNova U1 Unified Multimodal Model
SenseTime has officially released and open-sourced the SenseNova U1 series of models. Built on its proprietary NEO-unify…
Microsoft Open-Sources VibeVoice, a Cutting-Edge Speech AI Model
Microsoft has officially open-sourced the VibeVoice speech AI model, achieving frontier-level performance in speech synt…
Retro AI: 13B Language Model Talkie Trained Exclusively on Pre-1930 Texts
GPT co-creator Alec Radford's new project talkie has launched — a 13-billion-parameter language model trained on 260 bil…
Talkie: A 13-Billion-Parameter Retro Language Model Inspired by the 1930s
A 13-billion-parameter language model called Talkie has attracted widespread attention. Designed around the linguistic s…
Major Claude Pro Change: Opus Model Now Requires Extra Usage to Access
Anthropic has made a significant adjustment to its Claude Pro subscription plan. Its most powerful Opus model will no lo…
Microsoft Open-Sources VibeVoice Speech Recognition Model
Microsoft has released VibeVoice, an MIT-licensed speech-to-text model with built-in speaker diarization. Widely regarde…
MiniMax Releases M2.7 Model, Strengthening Agentic Workflows
MiniMax has officially released the M2.7 large language model, significantly enhancing agentic capabilities over M2.5, d…
NVIDIA Dynamo Full-Stack Optimization for Agent Inference Performance
NVIDIA has launched the Dynamo inference framework, delivering full-stack optimization for AI Agent workloads. As enterp…
NVIDIA Megatron Integrates Higher-Order Optimizers to Accelerate LLM Training
NVIDIA has integrated higher-order optimization algorithms such as Shampoo into its Megatron framework, significantly ac…
DeepSeek V4 Officially Launches Alongside NVIDIA Blackwell Support
DeepSeek releases its fourth-generation flagship models V4-Pro and V4-Flash, while NVIDIA provides GPU-accelerated endpo…