Trinity Large Preview
Open weights Preview Text LLM
MoE trinity-large-preview
Lightly post-trained instruct checkpoint of Arcee's 400B sparse MoE with 13B active parameters.
Released
Jan 27, 2026
Context window
524K tokens
Max output
Not disclosed
Parameters
398B 13B active
Input price
Not disclosed
Output price
Not disclosed
Prices are first-party list prices in USD at the standard tier. Last verified Oct 30, 2026.
Pricing
No list price recorded for this model yet.
Specifications
- Developer
- Arcee AI
- Model type
- Text LLM
- Architecture
- MoE
- Parameters
- 398B (13B active)
- Input
- Text
- Output
- Text
- Knowledge cutoff
- Not disclosed
- License
- Other (open)
- API access
- First-party API