Trinity-Large-Preview
Trinity-Large-Preview is a 398B-parameter sparse Mixture-of-Experts (MoE) model with approximately 13B active parameters per token.
text → text · made by Arcee AI
Sold by
Nobody in the catalogue publishes a price for this yet.
Measured 1 standing
| Place | Board | Metric | Score |
|---|---|---|---|
| 32ndof 111 | Vectara Hallucination Leaderboard | Hallucination Rate | 6.9 |
About
Trinity-Large-Preview — a text model from Arcee AI, placed 32nd of 111 on Vectara Hallucination Leaderboard.
It takes text and returns text, with a context window of 131,072 tokens. It was published in January 2026, trained on material up to June 2025. Its sellers say it can call a tool. The catalogue files it under extract. Its weights are published under the other licence, which attaches conditions the plain open licences do not. It stands 32nd of 111 on Vectara Hallucination Leaderboard.
Every current figure
- Maker
- Arcee AI
- Register
- model
- Takes
- text
- Returns
- text
- Context
- 131,072 tokens
- Published
- January 2026
- Knowledge to
- June 2025
- Licence
- other
- Boards
- 1
- Best place
- 32nd of 111 — Vectara Hallucination Leaderboard
Known as 1 name
arcee-ai/Trinity-Large-Preview