"On January 28th, a 30-person U.S. startup called Arcee AI released Trinity Large, a 400-billion-parameter sparse Mixture of Experts model that challenges Meta's Llama 4 Maverick and Chinese models like Z.ai's GLM-4.5. Trained in just six months for $20 million using 2,048 NVIDIA Blackwell B300 GPUs, Trinity represents a significant achievement in democratizing frontier-grade model development. What sets Trinity apart is its commitment to the Apache 2.0 license-a truly open-source alternative to Meta's proprietary Llama license."
"January 2026 kicked off with a clear message: AI is no longer just a tool, but a core component of the digital economy. The month was defined by the rapid integration of agentic AI into commerce, the launch of specialized AI workspaces for scientific research, and the continued explosion of open-source models - particularly from China. Meanwhile, a viral AI agent called Moltbot sparked both excitement and alarm about autonomous systems, while a tiny U.S. startup challenged Big Tech's dominance with a massive open-source model."
January 2026 showed AI as a central pillar of the digital economy, driven by agentic systems, specialized research workspaces, and surging open-source models from China and elsewhere. A viral agent named Moltbot highlighted both promise and concerns around autonomy. Arcee AI released Trinity Large, a 400B sparse Mixture of Experts model trained in six months on 2,048 Blackwell B300 GPUs for $20M, licensed under Apache 2.0, and delivering competitive performance across coding, math, reasoning, and knowledge benchmarks. Trinity’s Preview, Base, and TrueBase variants support developers, researchers, and enterprises. DeepSeek AI launched DeepSeek-OCR-2 as a state-of-the-art 3B vision-language document model.
Read at Scalac - Software Development Company - Akka, Kafka, Spark, ZIO
Unable to calculate read time
Collection
[
|
...
]