← all releases
Arcee AI·Jan 27, 2026·Preview, Base, TrueBase·open source
Trinity Large
Trinity Large is a 400-billion parameter sparse Mixture-of-Experts (MoE) model developed by Arcee AI. It utilizes 256 experts with 4 experts active per token, resulting in a sparsity ratio of 1.56%. The model was trained on 17 trillion tokens using 2,048 Nvidia B300 GPUs over a period of 33 days.
LLMMOE
overview
Trinity Large is a 400-billion parameter sparse Mixture-of-Experts (MoE) model developed by Arcee AI. It utilizes 256 experts with 4 experts active per token, resulting in a sparsity ratio of 1.56%. The model was trained on 17 trillion tokens using 2,048 Nvidia B300 GPUs over a period of 33 days.
key features
- 01400-billion parameter sparse MoE model
- 02256 experts with 4 experts active per token
- 03Sparsity ratio of 1.56%
- 04Trained on 17 trillion tokens
- 05Utilized 2,048 Nvidia B300 GPUs over 33 days
related productsbrowse all →
Protenixopen sourceProtenix is an open-source, trainable PyTorch implementation of AlphaFold 3, designed for high-accuracy biomolecular structure prediction. It aims to advance accessible and extensible research tools for the computational biology community. ([github.com](https://github.com/bytedance/Protenix?utm_source=openai))Z-Imageopen sourceZ-Image is an efficient image generation foundation model developed by Tongyi-MAI, designed to produce high-quality, diverse, and stylistically versatile images. It serves as a robust base for creators, researchers, and developers seeking advanced image generation capabilities.