MiroMind’s MiroThinker 1.5 delivers trillion-parameter performance from a 30B model — at 1/20th the cost

Joining the ranks of a growing number of smaller, powerful reasoning models is MiroThinker 1.5 from MiroMind, with just 30 ...
InfoQ

DeepSeek-V3.2 Outperforms GPT-5 on Reasoning Tasks

V3.2, a family of open-source reasoning and agentic AI models. The high compute version, DeepSeek-V3.2-Speciale, performs ...

Meta’s Vision-Language Shift VL-JEPA Beats Bulky LLMs

VL-JEPA predicts meaning in embeddings, not words, combining visual inputs with eight Llama 3.2 layers to give faster answers ...