Arcee AI

2026Arcee AIpaid

Arcee Open Models API Beta

vBeta
AI API
Open-Weight Models

The Arcee Open Models API Beta expands Arcee's hosted inference service beyond its proprietary Trinity model family to include a curated selection of frontier open-weight models. This service is designed to provide developers and enterprises with greater flexibility for demanding, long-horizon agentic workloads.

Access to a diverse catalog including Trinity-Large-Thinking, DeepSeek-V4-Pro, DeepSeek-V4-Flash-Latest, GLM-5.2, Kimi-K3, and Inkling-Small
Integration with 'nac', an open-source agent harness for long-running tasks
Token-based pricing model for flexible usage
versionBeta
regionUnited States
2026Arcee AIopen source

Trinity Large

LLM
MOE

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.

400-billion parameter sparse MoE model
256 experts with 4 experts active per token
Sparsity ratio of 1.56%
pricingFree on OpenRouter for a limited time
regionUnited States
2026Arcee AIopen source

NAC

v0.1.2
AI Agents
Agent Harness

NAC is an open-source agent harness designed to manage long-running, complex engineering tasks by separating temporary execution context from persistent workstream state. It utilizes a thread-and-episode architecture where a central orchestrator dispatches parallel workers to complete bounded tasks, ensuring agents maintain focus and intent over extended durations.

Thread-and-episode architecture for persistent state management
Central orchestrator for task decomposition and parallel worker dispatch
Thread weaving to synthesize information across multiple workstreams
version0.1.2
regionUnited States