Back to News
Arcee AIJanuary 27, 2026

Trinity Large

Open Source
LLM
MOE

Explore Trinity Large

Visit the official website to learn more and get started

### TL;DR

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 Insights & Metrics

Pricing
Free on OpenRouter for a limited time
Cost structure
Version
Preview, Base, TrueBase
Current release version
Hardware
2,048 Nvidia B300 GPUs
Compute requirements
Category
Open Source
Licensing model
Region
United States
Primary region

Key Features

  • 400-billion parameter sparse MoE model
  • 256 experts with 4 experts active per token
  • Sparsity ratio of 1.56%
  • Trained on 17 trillion tokens
  • Utilized 2,048 Nvidia B300 GPUs over 33 days

Discussion

0
Upvotes
0
Downvotes
0 reviews

Sign in to leave a review

Reviews

No reviews yet. Be the first to review!

🚀 Join the AI dev community — follow us everywhere

© 2026 MARKTECHPOST AI MEDIA INC. All rights reserved.Terms & ConditionsPrivacy Policy
Beta Mode