Trinity-Large-Thinking is a reasoning-optimized variant of Arcee AI's Trinity-Large family — a 398B-parameter sparse Mixture-of-Experts (MoE) model with approximately 13B active parameters per token, post-trained with extended chain-of-thought reasoning and agentic RL.
For full model details, benchmarks, and usage guidance, see the main
Trinity-Large-Thinking model card.
Supported in llama.cpp release b7061+.
1# Recommended quant
2llama-server -hf arcee-ai/Trinity-Large-Thinking-GGUF:Q4_K_M
3
4# Higher quality
5llama-server -hf arcee-ai/Trinity-Large-Thinking-GGUF:Q6_K
6
7# Lower memory
8llama-server -hf arcee-ai/Trinity-Large-Thinking-GGUF:Q3_K_M
Trinity-Large-Thinking-GGUF is released under the OpenMDW License, version 1.1 (OpenMDW-1.1).
1@misc{singh2026arceetrinity,
2 title = {Arcee Trinity Large Technical Report},
3 author = {Varun Singh and Lucas Krauss and Sami Jaghouar and Matej Sirovatka and Charles Goddard and Fares Obied and Jack Min Ong and Jannik Straube and Fern and Aria Harley and Conner Stewart and Colin Kealty and Maziyar Panahi and Simon Kirsten and Anushka Deshpande and Anneketh Vij and Arthur Bresnu and Pranav Veldurthi and Raghav Ravishankar and Hardik Bishnoi and DatologyAI Team and Arcee AI Team and Prime Intellect Team and Mark McQuade and Johannes Hagemann and Lucas Atkins},
4 year = {2026},
5 eprint = {2602.17004},
6 archivePrefix= {arXiv},
7 primaryClass = {cs.LG},
8 doi = {10.48550/arXiv.2602.17004},
9 url = {https://arxiv.org/abs/2602.17004}
10}
