Salience 27B R6 needs about 20.1 GB of memory at Q4_K_M (a 17.4 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 26 quants, from 0.0 GB to 54.7 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 8.7 GB.
Salience 27B R6 – a 27‑billion‑parameter vision‑language model with 1,048,576‑token context
Salience 27B R6 is a dense vision‑language model built for practical engineering tasks such as coding, debugging, and quantitative reasoning. It supports text and image input and includes a Multi‑Token Prediction head for speculative decoding. The model runs all 27B parameters on every token, providing maximum per‑token capacity.
Good at
- Coding and debugging real code
- Multi‑step terminal agency
- Quantitative reasoning
Limits
- No vision support without mmproj file
- Large memory requirement for full precision
- Requires llama.cpp release b10896 or newer
Best for
- Local LLM inference on high‑VRAM GPUs
- Agent loops needing fast stopping points
- Multimodal applications with image input
Who it’s for. Engineers and researchers who need a large, dense vision‑language model for complex coding and reasoning tasks. The model is suitable for local deployment on machines with sufficient VRAM and RAM.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.0 GB | 2.7 GB | Fits | Get file → | |
| IQ2_XXS | 8.9 GB | 11.6 GB | Too big | Get file → | |
| IQ2_XS | 9.1 GB | 11.8 GB | Too big | Get file → | |
| IQ2_S | 9.7 GB | 12.4 GB | Too big | Get file → | |
| IQ2_M | 10.5 GB | 13.2 GB | Too big | Get file → | |
| Q2_K | 10.8 GB | 13.5 GB | Too big | Get file → | |
| IQ3_XXS | 12.3 GB | 15.0 GB | Too big | Get file → | |
| Q3_K_S | 12.7 GB | 15.4 GB | Too big | Get file → | |
| IQ3_XS | 12.8 GB | 15.5 GB | Too big | Get file → | |
| Q3_K_M | 13.4 GB | 16.1 GB | Too big | Get file → | |
| Q3_K_L | 14.1 GB | 16.8 GB | Too big | Get file → | |
| IQ3_M | 14.9 GB | 17.5 GB | Too big | Get file → | |
| IQ4_XS | 15.5 GB | 18.2 GB | Too big | Get file → | |
| Q4_0 | 16.3 GB | 19.0 GB | Too big | Get file → | |
| Q4_K_S | 16.4 GB | 19.0 GB | Too big | Get file → | |
| IQ4_NL | 17.4 GB | 20.1 GB | Too big | Get file → | |
| Q4_K_M | 17.4 GB | 20.1 GB | Too big | Get file → | |
| Q4_1 | 17.8 GB | 20.5 GB | Too big | Get file → | |
| Q4_K_L | 18.8 GB | 21.5 GB | Too big | Get file → | |
| Q5_K_S | 19.6 GB | 22.2 GB | Too big | Get file → | |
| Q5_K_M | 20.9 GB | 23.6 GB | Too big | Get file → | |
| Q6_K_S | 22.9 GB | 25.5 GB | Too big | Get file → | |
| Q6_K | 23.9 GB | 26.5 GB | Too big | Get file → | |
| Q6_K_L | 25.0 GB | 27.6 GB | Too big | Get file → | |
| Q8_0 | 29.1 GB | 31.8 GB | Too big | Get file → | |
| BF16 | 54.7 GB (2 parts) | 57.3 GB | Too big | Get file → |
For 8 GB VRAM, use IQ2_XXS (8.88 GB file, 11.6 GB memory at 8 k context). For 12 GB VRAM, IQ2_S (9.68 GB file, 12.4 GB memory). For 16 GB VRAM, Q4_K_M (17.44 GB file, 20.1 GB memory). These estimates are approximate.
What GPU do you need for Salience 27B R6?
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 (0.0 GB file, about 2.7 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 2.7 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 2.7 GB in use).
- 12 GB VRAM: the best quant that fits is IQ2_XS (9.1 GB file, about 11.8 GB in use).
- 16 GB VRAM: the best quant that fits is IQ3_XS (12.8 GB file, about 15.5 GB in use).
- 24 GB VRAM: the best quant that fits is Q5_K_M (20.9 GB file, about 23.6 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (29.1 GB file, about 31.8 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (29.1 GB file, about 31.8 GB in use).
Which Salience 27B R6 quant should you choose?
Q4_K_M is the usual starting point for Salience 27B R6: it balances file size and quality well. Here it is a 17.4 GB file.
The Q8_0 file is 29.1 GB, about 1.7 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 0.0 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 Salience 27B R6
How much VRAM does Salience 27B R6 need?
About 20.1 GB at Q4_K_M with an 8K context window. That is the 17.4 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 Salience 27B R6 on an 8 GB GPU?
Yes, with the smaller files: the 0.0 GB file needs about 2.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Salience 27B R6 GGUF?
It depends on the quant: file is 0.0 GB, IQ2_XXS is 8.9 GB, IQ2_XS is 9.1 GB, IQ2_S is 9.7 GB, IQ2_M is 10.5 GB, Q2_K is 10.8 GB and 20 more.
What context length does Salience 27B R6 support?
The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 2.2 GB at 8K and 8.7 GB at 32K.
Where can I download Salience 27B R6 GGUF?
The files are on Hugging Face in the bartowski/vectionlabs_Salience-27B-R6-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 Salience 27B R6 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 Salience 27B R6 on 8 GB VRAM?
Use the IQ2_XXS quant, which is 8.88 GB file size and requires about 11.6 GB memory at 8 k context.
Does the model support image input?
Yes, but you need the mmproj projector files that pair with the chosen quant.
What is the context window size?
The model supports up to 1,048,576 tokens of context.