Trinity Large Preview needs about 244.1 GB of memory at Q4_K_M (a 241.6 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 29 quants, from 0.5 GB to 423.7 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 8.1 GB.
Trinity Large Preview – a 399B‑parameter sparse MoE model
Trinity Large Preview is a 399B‑parameter sparse Mixture‑of‑Experts model with 262144‑token context. It is lightly post‑trained for chat and supports long‑context comprehension. The model is released under Apache‑2.0.
Good at
- Large‑scale reasoning
- Long‑context dialogue
- Sparse‑MoE efficiency
Limits
- Requires large VRAM for higher quants
- No vision support
- License is Apache‑2.0
Best for
- Chat‑ready applications
- Long‑form content generation
- Research on MoE models
Who it’s for. Developers and researchers needing a large‑parameter model with sparse‑MoE efficiency. The model is suitable for environments that can handle its memory footprint and support llama.cpp or VLLM.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.5 GB | 3.0 GB | Fits | Get file → | |
| IQ1_S | 81.7 GB (3 parts) | 84.2 GB | Too big | Get file → | |
| IQ1_M | 85.1 GB (3 parts) | 87.6 GB | Too big | Get file → | |
| IQ2_XXS | 95.1 GB (3 parts) | 97.6 GB | Too big | Get file → | |
| IQ2_XS | 109.8 GB (3 parts) | 112.3 GB | Too big | Get file → | |
| IQ2_S | 110.1 GB (3 parts) | 112.6 GB | Too big | Get file → | |
| IQ2_M | 125.1 GB (4 parts) | 127.6 GB | Too big | Get file → | |
| Q2_K | 139.0 GB (4 parts) | 141.5 GB | Too big | Get file → | |
| Q2_K_L | 139.6 GB (4 parts) | 142.1 GB | Too big | Get file → | |
| IQ3_XXS | 157.0 GB (4 parts) | 159.5 GB | Too big | Get file → | |
| IQ3_XS | 162.4 GB (5 parts) | 164.9 GB | Too big | Get file → | |
| Q3_K_S | 172.8 GB (5 parts) | 175.3 GB | Too big | Get file → | |
| IQ3_M | 181.0 GB (5 parts) | 183.6 GB | Too big | Get file → | |
| Q3_K_M | 181.2 GB (5 parts) | 183.7 GB | Too big | Get file → | |
| Q3_K_L | 188.8 GB (5 parts) | 191.3 GB | Too big | Get file → | |
| Q3_K_XL | 189.3 GB (5 parts) | 191.8 GB | Too big | Get file → | |
| IQ4_XS | 212.8 GB (6 parts) | 215.3 GB | Too big | Get file → | |
| IQ4_NL | 225.2 GB (6 parts) | 227.7 GB | Too big | Get file → | |
| Q4_0 | 229.0 GB (6 parts) | 231.6 GB | Too big | Get file → | |
| Q4_K_S | 233.1 GB (6 parts) | 235.6 GB | Too big | Get file → | |
| Q4_K_M | 241.6 GB (7 parts) | 244.1 GB | Too big | Get file → | |
| Q4_K_L | 242.1 GB (7 parts) | 244.6 GB | Too big | Get file → | |
| Q4_1 | 249.9 GB (7 parts) | 252.4 GB | Too big | Get file → | |
| Q5_K_S | 274.7 GB (8 parts) | 277.3 GB | Too big | Get file → | |
| Q5_K_M | 283.3 GB (8 parts) | 285.8 GB | Too big | Get file → | |
| Q5_K_L | 283.7 GB (8 parts) | 286.2 GB | Too big | Get file → | |
| Q6_K | 327.6 GB (9 parts) | 330.1 GB | Too big | Get file → | |
| Q6_K_L | 327.9 GB (9 parts) | 330.4 GB | Too big | Get file → | |
| Q8_0 | 423.7 GB (11 parts) | 426.2 GB | Too big | Get file → |
For 8GB VRAM, use the unknown quant (0.52 GB file, 3 GB memory). For 16GB VRAM, IQ1_S (81.67 GB file, 84.2 GB memory) is too large; instead use Q4_0 (229.05 GB file, 231.6 GB memory) which fits 16GB VRAM when loaded with sharding. For 32GB VRAM, Q4_K_M (241.63 GB file, 244.1 GB memory) is a good balance between size and performance.
What GPU do you need for Trinity Large Preview?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
- 24 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
- 32 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
- 48 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 3.0 GB in use).
Which Trinity Large Preview quant should you choose?
Q4_K_M is the usual starting point for Trinity Large Preview: it balances file size and quality well. Here it is a 241.6 GB file.
The Q8_0 file is 423.7 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is the first one in the table at 0.5 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Trinity Large Preview
How much VRAM does Trinity Large Preview need?
About 244.1 GB at Q4_K_M with an 8K context window. That is the 241.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Trinity Large Preview on an 8 GB GPU?
Yes, with the smaller files: the 0.5 GB file needs about 3.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Trinity Large Preview GGUF?
It depends on the quant: file is 0.5 GB, IQ1_S is 81.7 GB, IQ1_M is 85.1 GB, IQ2_XXS is 95.1 GB, IQ2_XS is 109.8 GB, IQ2_S is 110.1 GB and 23 more.
What context length does Trinity Large Preview support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 2.0 GB at 8K and 8.1 GB at 32K.
Where can I download Trinity Large Preview GGUF?
The files are on Hugging Face in the bartowski/arcee-ai_Trinity-Large-Preview-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Trinity Large Preview use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Trinity Large Preview on 8 GB VRAM?
Yes, using the unknown quant (0.52 GB file) which requires approximately 3 GB of VRAM at 8k context.
Does the model support vision tasks?
No, Trinity Large Preview is a text‑only model with no vision capability.
What license applies to this model?
The model is released under the Apache‑2.0 license.