Skip to content
Log In
Sign Up
Dashboard

Trinity Large Preview

Trinity Large Preview is a 400B-parameter sparse mixture-of-experts model from Arcee AI that activates 13B parameters per forward pass, targeting math, coding, and multi-step agent workloads across a context window of 131K tokens.

Tool Use
index.ts
import { streamText } from 'ai'


const result = streamText({
  model: 'arcee-ai/trinity-large-preview',
  prompt: 'Why is the sky blue?'
})
OverviewPlaygroundAboutProvidersUptimeStatusSimilarFAQ

About Trinity Large Preview

Trinity Large Preview is a sparse mixture-of-experts model with 400B total parameters and 13B active per forward pass. The MoE design keeps per-token compute proportional to the active parameter count while the full parameter space retains broad knowledge.

Arcee AI targets math, coding, and multi-step agent workflows with this release. Extended multi-turn sessions stay efficient because only the active expert subset runs on each token.

This release is labeled preview, so treat behavior, pricing, and versioning as subject to change while you benchmark. Use https://docs.arcee.ai/language-models/trinity-large-400b for the latest specs and rates from Arcee AI.