Arcee AI Trinity Large 400B Architecture
Back to Arcee AI Trinity Large 400B in the gallery
Arcee AI
Arcee's flagship blends several efficiency tricks into a DeepSeek-like coarse MoE design; the AA score shown here refers to the Trinity Large Thinking checkpoint.
Architecture and specifications
- Scale
- 400B total, 13B active (3.3% active)
- Context (tokens)
- 512,000
- License
- Apache License 2.0
- Date
- 2026-01-27
- Decoder type
- Sparse MoE
- Attention
- GQA with gated attention and 3:1 sliding-window/global attention
- Layer mix
- 45 sliding-window + 15 global
- KV cache / token (bf16) info
- 240 KiB · High
- Key detail
- Combines QK-Norm, RoPE+NoPE, sandwich norm, and a coarse-grained MoE.
Training and benchmarks
- Pretraining data
- 17T tokens Full Base checkpoint, including its mid-training anneals; the earlier TrueBase checkpoint used 10T. Trinity Large report · Trinity Large Thinking model card
- Post-training methods
- SFT, RL Preview uses SFT; the later Thinking checkpoint adds instruction tuning and agentic RL. Trinity Large report · Trinity Large Thinking model card