Model reference · open weights

Trinity-Large-True

Available as managed deployment Licence fee LLMs arcee-ai Text gen 1 variants 234 dl/mo

Trinity-Large-True is an open-weight language model from arcee-ai. AxForge deploys and operates it for you on dedicated EU-owned hardware — with the licence handled where one is required.

Available as managed deployment — configured and operated for you on dedicated EU hardware, quoted per deployment.

What it is

Makerarcee-ai
TypeLanguage models
TaskText gen
Parameters (lead)398.6B
Context8k tokens
Runs withtransformers
Released2026-01-27
Popularity234 downloads / month
LicenceCommercial licence needed

About

What Trinity-Large-True is

src="https://cdn-uploads.huggingface.co/production/uploads/6435718aaaef013d1aec3b8b/i-v1KyAMOW_mgVGeic9WJ.png" alt="Arcee Trinity Large" style="max-width: 100%; height: auto;" >

Trinity-Large-TrueBase

Introduction

Trinity-Large-TrueBase is a base pretraining checkpoint from Arcee AI's Trinity Large training run. It is a 398B-parameter sparse Mixture-of-Experts (MoE) model with approximately 13B active parameters per token. The checkpoint was captured after 10 trillion tokens of pretraining, prior to learning-rate annealing and before any instruction tuning or reinforcement learning.

This checkpoint is intended for research, probing, ablation studies, and downstream fine-tuning and comes without any pre-baked alignment, instruction formatting, or preference optimization.

More details on the training of Trinity Large are available in the technical report.

Model Variants

The Trinity Large family consists of three checkpoints from the same training run:

  • Trinity-Large-TrueBase (this release): 10T-token pre-anneal checkpoint with no instruction data
  • Trinity-Large-Thinking: Reasoning-optimized, agentic post-training with extended chain-of-thought
  • Trinity-Large-Base: Full 17T-token pretrained foundation model with mid-training anneals
  • Trinity-Large-Preview: Lightly post-trained, chat-ready model undergoing active RL

Architecture

Trinity-Large-TrueBase uses a sparse MoE configuration designed to maximize efficiency while maintaining large-scale capacity.

HyperparameterValue
Total parameters~398B
Active parameters per token~13B
Experts256
Active experts4
Routing strategy4-of-256 (1.56% sparsity)
Dense layers6
Pretraining context length8,192
ArchitectureSparse MoE (AfmoeForCausalLM)

Note: Extended context support (e.g., 512k) was introduced after this checkpoint and is not available in TrueBase.

Benchmark Results

BenchmarkN-shotMetricScoreStderr
arc_challenge_0shot0acc_norm,none0.6237±0.0142
bbh_fewshot3exact_match,remove_whitespace0.5784±0.0054
gpqa_diamond_5shot5acc_norm,none0.4091±0.0350
gpqa_diamond_generative_5shot5exact_match,flexible-extract0.3788±0.0346
gsm8k_8shot8exact_match,flexible-extract0.8036±0.0109
gsm8k_cot8exact_match,flexible-extract0.8044±0.0109
hellaswag_5shot5acc_norm,none0.8813±0.0032
humaneval_plus0pass@1,create_test0.5183±0.0391
leaderboard_math_hard4exact_match,none0.2696±0.0113
mbpp_plus3pass_at_1,none0.8095±0.0202
minerva_math5004math_verify,none0.4820±0.0224
mmlu_5shot5acc,none0.7845±0.0033
mmlu_generative_5shot5exact_match,get_response0.7848±0.0033
mmlu_pro5exact_match,custom-extract0.5160±0.0044
triviaqa_5shot5exact_match,remove_whitespace0.8096±0.0029
winogrande_5shot5acc,none0.8145±0.0109

Training Configuration

Pretraining

  • Training tokens: 10 trillion
  • Checkpoint type: Pre-anneal
  • Instruction data: None
  • RLHF or post-training: None

This checkpoint branches from the main Trinity Large run at the 10T-token mark, prior to learning-rate decay or post-training phases.

Optimizers

Optimizer learning rates after WSD warm-up:

  • Adam learning rate: 2e-4
  • Muon learning rate: 8e-4

Muon was used to support larger critical batch sizes in a highly sparse MoE regime.

Infrastructure

  • Hardware: 2,048 NVIDIA B300 GPUs
  • Parallelism: HSDP + Expert Parallelism
  • Compute partner: Prime Intellect
  • Data partner: Datology

Intended Use

  • Studying emergent behavior from large-scale pretraining
  • Sparse MoE routing and load-balancing research
  • Interpretability, probing, and ablation studies
  • Domain-specific fine-tuning from a clean base
  • Academic and industrial foundation model research

Rationale for Release

Most base model releases include instruction data, annealed training dynamics, or early alignment stages. Trinity-Large-TrueBase excludes these, providing an opportunity to study what large-scale models learn from pretraining data alone. This checkpoint is intended as a foundation for research rather than as a finished conversational assistant.

Known Limitations

  • Not aligned for safety, helpfulness, or conversational tone
  • Requires substantial compute and expertise to fine-tune
  • May exhibit raw or unstable behaviors typical of unaligned models
  • No extended-context tuning beyond the 8K pretraining window

License

Trinity-Large-TrueBase is released under the OpenMDW License, version 1.1 (OpenMDW-1.1).

From the published model card. Full card on the HuggingFace links in the sidebar.

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys trinity-large-true for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (trinity-large-true below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/chat/completions \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"trinity-large-true","messages":[{"role":"user","content":"Hello"}]}'

Create an account — your API key is available in the console. 5M tokens/month currently included with every new account at launch.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms