Total runs:
24.5K
Run Growth:
491
Growth Rate:
2.01%
Sponsored by
APIMart
arcee-ai / Trinity-Large-Preview-FP8
Model's Last Updated: January 28 2026
text-generation
Introduction of Trinity-Large-Preview-FP8
Model Details of Trinity-Large-Preview-FP8
Trinity-Large-Preview-FP8
Introduction
Trinity-Large-Preview is a 398B-parameter sparse Mixture-of-Experts (MoE) model with approximately 13B active parameters per token. It is the largest model in Arcee AI's Trinity family, trained on more than 17 trillion tokens and delivering frontier-level performance with strong long-context comprehension. Trinity-Large-Preview is a lightly post-trained model based on Trinity-Large-Base.
This repository contains the FP8 quantized weights of Trinity-Large-Preview.
Try it at chat.arcee.ai
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-Preview : Lightly post-trained, chat-ready model undergoing active RL
- Trinity-Large-TrueBase : 10T-token pre-anneal pretraining checkpoint
- Trinity-Large-Base : Full 17T-token pretrained foundation model with mid-training anneals
Architecture
Trinity-Large-Preview uses a sparse MoE configuration designed to maximize efficiency while maintaining large-scale capacity.
| Hyperparameter | Value |
|---|---|
| Total parameters | ~398B |
| Active parameters per token | ~13B |
| Experts | 256 (1 shared) |
| Active experts | 4 |
| Routing strategy | 4-of-256 (1.56% sparsity) |
| Dense layers | 6 |
| Pretraining context length | 8,192 |
| Context length after extension | 512k |
| Architecture | Sparse MoE (AfmoeForCausalLM) |
Benchmarks
| Benchmark | Llama 4 Maverick | Trinity-Large Preview |
|---|---|---|
| MMLU | 85.5 | 87.2 |
| MMLU-Pro | 80.5 | 75.2 |
| GPQA-Diamond | 69.8 | 63.3 |
| AIME 2025 | 19.3 | 24.0 |
Training Configuration
Pretraining
- Training tokens: 17 trillion
- Data partner: Datology
Posttraining
- This checkpoint was instruction tuned on 20B tokens.
Infrastructure
- Hardware: 2,048 NVIDIA B300 GPUs
- Parallelism: HSDP + Expert Parallelism
- Compute partner: Prime Intellect
Usage
Running our model
Recommended settings
- temperature:
- top_k:
- top_p:
- min_p:
Transformers
Use the
main
transformers branch or pass
trust_remote_code=True
with a released version.
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
model_id = "arcee-ai/Trinity-Large-Preview-FP8"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
model_id,
device_map="auto",
trust_remote_code=True
)
messages = [
{"role": "user", "content": "Who are you?"},
]
input_ids = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt"
).to(model.device)
outputs = model.generate(
input_ids,
max_new_tokens=256,
do_sample=True,
temperature=0.8,
top_k=50,
top_p=0.8
)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)
VLLM
Supported in VLLM release 0.11.1+
vllm serve arcee-ai/Trinity-Large-Preview-FP8 \
--enable-auto-tool-choice \
--tool-call-parser hermes
API
Available on OpenRouter:
curl -X POST "https://openrouter.ai/v1/chat/completions" \
-H "Authorization: Bearer $OPENROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "arcee-ai/trinity-large-preview",
"messages": [
{
"role": "user",
"content": "What are some fun things to do in New York?"
}
]
}'
License
Trinity-Large-Preview is released under the Apache License, Version 2.0.
Citation
@misc{arcee_trinity_large_preview,
title = {Trinity-Large-Preview},
author = {{Arcee AI}},
year = {2026},
note = {398B sparse MoE model trained on 17T tokens}
}
Runs of arcee-ai Trinity-Large-Preview-FP8 on huggingface.co
614
Total runs
0
24-hour runs
-2.8K
3-day runs
-4.4K
7-day runs
-4.4K
30-day runs
More Information About Trinity-Large-Preview-FP8 huggingface.co Model
Url of Trinity-Large-Preview-FP8
Trinity-Large-Preview-FP8 huggingface.co Url
Provider of Trinity-Large-Preview-FP8 huggingface.co
Other API from arcee-ai
Updated:May 29 2026
Total runs:
14.3K
Run Growth:
14.3K
Growth Rate:
99.67%
Total runs:
4.2K
Run Growth:
-3.1K
Growth Rate:
-72.82%
Total runs:
2.8K
Run Growth:
2.2K
Growth Rate:
78.46%
Total runs:
2.5K
Run Growth:
1.2K
Growth Rate:
49.37%
Total runs:
1.9K
Run Growth:
1.0K
Growth Rate:
56.52%
Total runs:
1.8K
Run Growth:
1.5K
Growth Rate:
82.87%
Total runs:
1.7K
Run Growth:
-860
Growth Rate:
-49.60%
Total runs:
1.7K
Run Growth:
1.6K
Growth Rate:
94.76%
Total runs:
1.6K
Run Growth:
-3.9K
Growth Rate:
-245.51%
Total runs:
1.5K
Run Growth:
1.3K
Growth Rate:
86.22%
Total runs:
1.5K
Run Growth:
1.5K
Growth Rate:
98.63%
Total runs:
1.5K
Run Growth:
1.3K
Growth Rate:
86.79%
Total runs:
1.4K
Run Growth:
504
Growth Rate:
34.95%
Total runs:
1.3K
Run Growth:
-107
Growth Rate:
-8.24%
Total runs:
1.2K
Run Growth:
-106
Growth Rate:
-8.55%
Total runs:
1.2K
Run Growth:
1.1K
Growth Rate:
89.11%
Total runs:
1.1K
Run Growth:
80
Growth Rate:
7.04%
Total runs:
1.0K
Run Growth:
-6
Growth Rate:
-0.58%
Total runs:
1.0K
Run Growth:
-1.8K
Growth Rate:
-174.58%
Total runs:
850
Run Growth:
0
Growth Rate:
0.00%
Total runs:
766
Run Growth:
-6.3K
Growth Rate:
-820.89%
Total runs:
765
Run Growth:
692
Growth Rate:
90.46%
Total runs:
537
Run Growth:
140
Growth Rate:
26.07%
Total runs:
528
Run Growth:
-163
Growth Rate:
-30.87%
Total runs:
505
Run Growth:
201
Growth Rate:
39.80%
Total runs:
503
Run Growth:
382
Growth Rate:
75.94%
Total runs:
471
Run Growth:
323
Growth Rate:
68.58%
Total runs:
455
Run Growth:
417
Growth Rate:
91.65%
Total runs:
448
Run Growth:
107
Growth Rate:
23.88%
Total runs:
410
Run Growth:
175
Growth Rate:
42.68%
Total runs:
408
Run Growth:
0
Growth Rate:
0.00%
Total runs:
351
Run Growth:
351
Growth Rate:
100.00%
Total runs:
281
Run Growth:
0
Growth Rate:
0.00%
Total runs:
246
Run Growth:
143
Growth Rate:
58.13%
Total runs:
243
Run Growth:
-1.7K
Growth Rate:
-680.25%
Total runs:
192
Run Growth:
111
Growth Rate:
57.81%
Total runs:
182
Run Growth:
108
Growth Rate:
59.34%
Total runs:
178
Run Growth:
71
Growth Rate:
39.89%
Total runs:
174
Run Growth:
62
Growth Rate:
36.26%
Total runs:
158
Run Growth:
-1
Growth Rate:
-0.63%
Total runs:
152
Run Growth:
98
Growth Rate:
64.47%
Total runs:
148
Run Growth:
-99
Growth Rate:
-66.89%
Total runs:
139
Run Growth:
61
Growth Rate:
43.88%
Total runs:
137
Run Growth:
74
Growth Rate:
54.01%
Total runs:
136
Run Growth:
40
Growth Rate:
29.41%
Total runs:
130
Run Growth:
75
Growth Rate:
59.52%
Total runs:
125
Run Growth:
65
Growth Rate:
52.00%
Total runs:
122
Run Growth:
74
Growth Rate:
60.66%
Total runs:
120
Run Growth:
70
Growth Rate:
58.33%
Total runs:
119
Run Growth:
67
Growth Rate:
57.26%
Total runs:
119
Run Growth:
71
Growth Rate:
59.66%
Total runs:
120
Run Growth:
69
Growth Rate:
58.47%
Total runs:
118
Run Growth:
67
Growth Rate:
56.78%
Total runs:
119
Run Growth:
59
Growth Rate:
50.43%
Total runs:
114
Run Growth:
65
Growth Rate:
57.02%
Total runs:
113
Run Growth:
61
Growth Rate:
54.95%
Total runs:
113
Run Growth:
62
Growth Rate:
55.86%
Total runs:
113
Run Growth:
66
Growth Rate:
58.41%
Total runs:
112
Run Growth:
61
Growth Rate:
54.46%
Total runs:
110
Run Growth:
62
Growth Rate:
56.36%
Total runs:
110
Run Growth:
58
Growth Rate:
54.21%
Total runs:
109
Run Growth:
60
Growth Rate:
56.07%
Total runs:
109
Run Growth:
58
Growth Rate:
53.21%
Total runs:
107
Run Growth:
65
Growth Rate:
61.90%
Total runs:
107
Run Growth:
63
Growth Rate:
58.88%
Total runs:
106
Run Growth:
53
Growth Rate:
50.00%
Total runs:
104
Run Growth:
53
Growth Rate:
51.96%
Total runs:
104
Run Growth:
55
Growth Rate:
52.88%
Total runs:
104
Run Growth:
53
Growth Rate:
50.96%
Total runs:
103
Run Growth:
55
Growth Rate:
53.40%
Total runs:
102
Run Growth:
49
Growth Rate:
48.04%
Total runs:
100
Run Growth:
54
Growth Rate:
54.00%